UBC Community, Partners, and Alumni Publications

INCOFISH ecosystem models: Transiting from Ecopath to Ecospace. Le Quesne, Will J.F.; Arreguín-Sánchez, Francisco; Heymans, Sheila J.J. 2007

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

Item Metadata

Download

Media
[if-you-see-this-DO-NOT-CLICK]
[if-you-see-this-DO-NOT-CLICK]
FCRR_2007_15-6.pdf [ 14.12MB ]
[if-you-see-this-DO-NOT-CLICK]
Metadata
JSON: 1.0074767.json
JSON-LD: 1.0074767+ld.json
RDF/XML (Pretty): 1.0074767.xml
RDF/JSON: 1.0074767+rdf.json
Turtle: 1.0074767+rdf-turtle.txt
N-Triples: 1.0074767+rdf-ntriples.txt
Original Record: 1.0074767 +original-record.json
Full Text
1.0074767.txt
Citation
1.0074767.ris

Full Text

ISSN 1198-6727 INCOFISH ECOSYSTEM MODELS: TRANSITING FROM ECOPATH TO ECOSPACE  Fisheries Centre Research Reports 2007 Volume 15 Number 6  ISSN 1198-6727  Fisheries Centre Research Reports 2007 Volume 15 Number 6  INCOFISH Ecosystem Models: Transiting from Ecopath to Ecospace  Fisheries Centre, University of British Columbia, Canada  INCOFISH Ecosystem Models: Transiting from Ecopath to Ecospace  edited by Will J.F. Le Quesne, Francisco Arreguín-Sánchez and Sheila J.J. Heymans  Fisheries Centre Research Reports 15 (6) 188 pages © published 2007 by The Fisheries Centre University of British Columbia 2202 Main Mall Vancouver, B.C., Canada, V6T 1Z4  ISSN 1198-6727  Fisheries Centre Research Reports 15 (6) 2007 INCOFISH ECOSYSTEM MODELS: TRANSITING FROM ECOPATH TO ECOSPACE edited by  Will J.F. Le Quesne, Francisco Arreguín-Sánchez and Sheila J.J. Heymans  CONTENTS Page DIRECTOR’S FOREWORD ...................................................................................................................................... 1 PREFACE .............................................................................................................................................................2 TROPHIC MODEL OF THE NORTHERN ADRIATIC SEA, AN EUTROPHIC AND HIGHLY EXPLOITED ECOSYSTEM, A. Barausse, A. Duci, C. Mazzoldi, Y. Artioli and L. Palmeri..................................................................3 Abstract .....................................................................................................................................................3 Introduction ..............................................................................................................................................3 Materials and Methods .............................................................................................................................3 Results .....................................................................................................................................................10 Discussion................................................................................................................................................ 11 Acknowledgements ................................................................................................................................. 12 References ............................................................................................................................................... 12 Appendix A .............................................................................................................................................. 16 UPDATED ECOSYSTEM MODEL FOR THE NORTHERN BENGUELA ECOSYSTEM, NAMIBIA, S.J.J. Heymans and U.R. Sumaila ...........................................................................................................................................25 Abstract ...................................................................................................................................................25 Introduction ............................................................................................................................................25 Materials and Methods ...........................................................................................................................26 Model Fitting Results ..............................................................................................................................57 Acknowledgements .................................................................................................................................63 References ...............................................................................................................................................63 Appendix A ............................................................................................................................................. 69 MODELLING THE FOOD WEB IN THE UPWELLING ECOSYSTEM OFF CENTRAL CHILE (33°S–39°S) IN THE YEAR 2000, S. Neira and H. Arancibia................................................................................................... 71 Abstract ................................................................................................................................................... 71 Introduction ............................................................................................................................................ 71 Methods...................................................................................................................................................72 References .............................................................................................................................................. 83 SPATIAL RESOURCES AND FISHERY MANAGEMENT FRAMEWORK IN THE EAST CHINA SEA, H.Q. Cheng, H. Jiang, H.G. Xu, J. Wu, H. Ding, W. Le Quesne and F. Arreguín-Sánchez ......................................87 Abstract ...................................................................................................................................................87 Introduction ............................................................................................................................................87 Spatial Characteristics of the ECS Ecosystem ....................................................................................... 88 Spatial Modelling of Fishery Management Scenarios in the ECS ..........................................................92 Acknowledgements ................................................................................................................................ 96 References .............................................................................................................................................. 96 Appendix A ............................................................................................................................................. 98  AN ECOSYSTEM SIMULATION MODEL OF THE NORTHERN GULF OF CALIFORNIA, D. Lercari, F. ArreguínSánchez and W. Le Quesne .................................................................................................................. 100 Abstract................................................................................................................................................. 100 Introduction.......................................................................................................................................... 100 Description of Model Components: Ecopath ....................................................................................... 102 Description of Model Components: Ecosim ........................................................................................ 105 Description of Model Components: Ecospace ..................................................................................... 108 Conclusions............................................................................................................................................ 111 References.............................................................................................................................................. 112 MARINE ECOSYSTEM ANALYSES IN THE GULF OF ULLOA, MEXICO: BAC MEETS ECOPATH, P. del Monte-Luna, F. Arreguín-Sánchez and D. Lluch-Belda............................................................................................. 114 Abstract.................................................................................................................................................. 114 Introduction........................................................................................................................................... 114 Input Data and Ecosystem Characterization ........................................................................................118 The Ecosim Foundation......................................................................................................................... 119 The Ecospace Foundation ..................................................................................................................... 121 Results....................................................................................................................................................123 Discussion ..............................................................................................................................................127 Acknowledgements................................................................................................................................129 References..............................................................................................................................................129 Appendix A.............................................................................................................................................133 TROPHIC MODEL FOR THE ECOSYSTEM OF LA PAZ BAY, SOUTHERN BAJA CALIFORNIA PENINSULA, MEXICO, F. Arreguín-Sánchez, P. del Monte-Luna, J.G. Díaz-Uribe, M. Gorostieta, E.A. Chávez and R. Ronzón-Rodríguez.................................................................................................................................134 Abstract..................................................................................................................................................134 Introduction...........................................................................................................................................134 Materials and Methods..........................................................................................................................135 Results....................................................................................................................................................137 Discussion ..............................................................................................................................................147 Acknowledgements............................................................................................................................... 150 References............................................................................................................................................. 150 Appendices.............................................................................................................................................153 SPATIAL MODELLING OF THE SENEGAMBIAN ECOSYSTEM, B. Samb.................................................................. 161 Abstract.................................................................................................................................................. 161 Introduction........................................................................................................................................... 161 Ecopath Model Development ................................................................................................................ 161 Ecospace Model Development ..............................................................................................................165 Conclusions............................................................................................................................................167 References............................................................................................................................................. 168 A BENTHIC ECOSYSTEM MODEL OF THE SINALOA CONTINENTAL SHELF, MEXICO, L.A. Salcido-Guevara and F. Arreguín-Sánchez .............................................................................................................................170 Abstract..................................................................................................................................................170 Introduction...........................................................................................................................................170 Materials and Methods.......................................................................................................................... 171 Model Balancing .................................................................................................................................... 176 Discussion............................................................................................................................................. 180 Acknowledgements................................................................................................................................182 References .............................................................................................................................................182 Appendix A ............................................................................................................................................187 A Research Report from the Fisheries Centre at UBC 188 pages © Fisheries Centre, University of British Columbia, 2007  FISHERIES CENTRE RESEARCH REPORTS ARE ABSTRACTED IN THE FAO AQUATIC SCIENCES AND FISHERIES ABSTRACTS (ASFA)  ISSN 1198-6727  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  1  DIRECTOR’S FOREWORD  Once again a Fisheries Centre Research Report was produced which features an international cast of authors, and documenting a number of ecosystem models constructed using the Ecopath with Ecosim (EwE) suite of software.  This report, however, also very nicely illustrates the increased sophistication of the users of this software, which parallels the improvements of EwE. Thus, we have here a number of models which emphasize the spatial dimension of EwE, i.e., the use of its Ecospace module. While technically not difficult to use, this module implies familiarity with the spatial dimension of an ecosystem. This is a dimension that all ecosystem modelers should be expected to master, though it may not be apparent to casual users of Ecopath, used for describing food webs, and Ecosim, used to simulate how they may change through time.  This report also illustrates another aspect of ecosystem modeling based on EwE, i.e., that numerous initiatives centered on this modeling approach are emerging outside of UBC's Fisheries Centre. In this case, the initiative was the INCOFISH Project (see www.incofish.org), funded by the European Commission, and which gathered a vast number of international collaborators, only some of whom are associated with the Fisheries Centre. Still, our Villy Christensen interacted with most of the authors of contributions included herein, one of the reasons for their quality.  Finally, I want to congratulate the editors and the authors of the contributions in this report for the enormous amount of contextualized ecological data that they are herewith making available to colleagues, and thus advancing ecosystem modeling everywhere.  Daniel Pauly Director, Fisheries Centre  2  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  PREFACE The INCOFISH project (“Integrating multiple demands on coastal zones with emphasis on fisheries and aquatic ecosystems”, European Commission contract 003739) was developed to examine integrated coastal zone management with explicit recognition of ecological, economic and social interactions due to the growing need for ecosystem-based management. A significant component of the overall INCOFISH project is based upon analysis of the ecosystem effects of management with the Ecopath modelling framework. The INCOFISH project is a global study. Models developed within this project include ecosystems from all five major continents. Specifically, the models include the East China Sea, Gulf of California, Humboldt Current, Northern Benguela, Gulf of Mexico, North Adriatic Sea, West Coast of Baja California, and Senegambian coasts. The Ecopath modelling framework (www.ecopath.org) includes Ecopath, Ecosim and Ecospace modules. Ecopath develops a static mass balanced network of interactions between components that make up an ecosystem. Ecosim builds upon this to allow time dynamic simulation of ecosystem interactions. Ecosim can incorporate environmental drivers and calibration to independent time series. Ecospace then builds on Ecosim to allow the spatio-temporal analysis of the ecosystem. It includes essential habitats and dispersion for functional groups, and incorporates socio-economic data to attempt to provide realistic simulations of spatial fishery interactions. The temporal (Ecosim) and spatial (Ecospace) components can be used as policy exploration tools. Critical to the successful use of the Ecopath suite is detailed and accurate parameterization of the Ecopath, Ecosim and Ecospace models for the study region selected. It is essential to document the development procedure to allow critical analysis of any studies subsequently based upon these models. This report provides detailed descriptions of the Ecopath, Ecosim and Ecospace models developed within the INCOFISH project. Some reports describe new and previously undescribed Ecopath models, while others build on previously described Ecopath models and therefore include mainly Ecosim and/or Ecospace components. We acknowledge the efforts of the authors contributing to this report. The Editors  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  3  TROPHIC MODEL OF THE NORTHERN ADRIATIC SEA, AN EUTROPHIC AND HIGHLY EXPLOITED ECOSYSTEM1 Alberto Baraussea,c, Alessandro Ducib, Carlotta Mazzoldib, Yuri Artiolia and Luca Palmeria a-  LASA, University of Padova, via Marzolo 9, 35131, Padova, Italy b - Dept. of Biology, University of Padova, via U. Bassi 58/B, 35121, Padova, Italy c - Email: alberto.barausse@unipd.it  ABSTRACT A preliminary Ecopath model of the Northern Adriatic Sea (Italy, Slovenia, Croatia) was developed. The model describes trophic fluxes in a marine ecosystem that is strongly impacted by human actions, being highly fished and characterised by eutrophication due to both anthropogenic and natural causes.  INTRODUCTION The Northern Adriatic Sea is a shallow, semi-enclosed basin of the Mediterranean Sea and surrounded by Italy, Slovenia and Croatia. It is characterised by strong inter-annual variability in the circulation field connected with meteorological conditions and with huge and fluctuating Po River freshwater input (Russo and Artegiani, 1996; Oddo et al., 2005). The Po River is also responsible for bringing a high load of nutrients into the basin, which, together with the intrinsic shallowness of the Northern Adriatic Sea, causes eutrophication. Temperature shows important seasonal oscillations, while salinity is connected with river runoff (Russo and Artegiani, 1996). The Northern Adriatic Sea ecosystem is subjected to strong human influence, mainly eutrophication and fisheries, as it is one of the most fished Italian seas. A preliminary model of the Northern Adriatic marine ecosystem trophic network was developed using the Ecopath software package. The data sources and calculations used to parametrize this model are presented in this report.  MATERIALS AND METHODS Ecopath modelling approach Ecopath (www.ecopath.org) is a well-known and accepted method for modelling aquatic ecosystems. In an Ecopath model, organisms are arranged together in functional groups (boxes) defined according to some criteria (e.g. taxonomy or common trophic features) and trophic interactions are represented as fluxes from one box to another. Fishing is also accounted for as a flux from the exploited box to the outside of the ecosystem. Ecopath food web models are based on the assumption of steady state and are, consequently, time averaged for a selected period. Hence, an Ecopath static network is an ‘instant snapshot’ of biomass and energy fluxes flowing in the ecosystem in the modelled time span. A short description of Ecopath follows, but a more comprehensive treatment can be found in Christensen and Walters (2004) and in Christensen et al. (2005). Data input demand for such models is relatively simple, as each functional group only requires information on diet (a diet matrix DC is used, whose elements are the percentage of how much a group 1 Cite as: Barausse, A., Duci, A., Mazzoldi, C., Artioli, Y. and Palmeri, L. 2007. Trophic model of the Northern Adriatic Sea, an eutrophic and highly exploited ecosystem, p. 3–24. In: Le Quesne, W.J.F., Arreguín-Sánchez, F. and Heymans, S.J.J. (eds.) INCOFISH ecosystem models: transiting from Ecopath to Ecospace. Fisheries Centre Research Reports 15(6). Fisheries Centre, University of British Columbia [ISSN 1198-6727].  4  Trophic Model of the Northern Adriatic Sea, Barausse, Duci, Mazzoldi, Artioli and Palmeri  contributes to another’s diet), biomass (B, e.g. expressed in wet weight, but also in energy currencies), production rate (P/B, equal to total mortality Z=M+F (sum of fishing and natural mortality) in steady state assumption), consumption rate (Q/B, the quantity of food consumed relative to biomass), catch (Y, accounting for discards and unreported catch), percentage of unassimilated consumption (GS, excreted and egested food percentage), plus, if any, net emigration flux (E) and biomass accumulation (BA) for a group in the simulated time period. All inputs should refer to the time period under consideration. Ecopath is based on two main equations derived from the steady-state assumption. The mass balance of fluxes entering and leaving a generic compartment i can be written: production = predation mortality + catches + other mortality + net emigration + biomass accumulation or n  Bi ⋅ ( P / B ) i − ∑ B j ⋅ (Q / B ) j ⋅ DC ji − ( P / B ) i ⋅ Bi ⋅ (1 − EEi ) − Yi − Ei − BAi = 0  Eq. 1  i =1  where n is the number of groups in the system, and DCji represents the percentage that i group constitutes in the diet of j group. EE is called ecotrophic efficiency and is the fraction of the production of a group that is used in the system. Consequently, (1-EE)·P/B is the so-called ‘other mortality’ due to disease or old age. Repeating the equation for every group of the model, a linear system can be written, with unknowns B, P/B, Q/B and EE (since C, GS, E, BA are parameters). So, the modeller must specify three out of four of these unknowns in every equation for Ecopath to solve the system and calculate the other unknowns. Usually, B, P/B and Q/B are entered, because EE is impossible to calculate experimentally and can only be guessed. However, in case of great uncertainty even on the order of magnitude of one of the three other unknowns for a group, EE can reasonably be entered to make the system possible to solve. Also, an energy balance can be written for every compartment: consumption = production + respiration + unassimilated food or  Ri = (1 − GS i ) ⋅ Qi − Pi  Eq. 2  where R is respiration. Usually R is calculated from this equation, but if respiration values are known it can be also used for calibrating GS or other input data in order to achieve the desired R value. It is worth noticing that an Ecopath model cannot be calibrated since accurate EE (and often also R) values for functional groups are not available. So, the mass and energy balance equations above are very useful in imposing a very simple, but powerful constraint on the input data: the model can be considered ‘done’ only when it is balanced, and mass and energy conservation are assured. These criteria are met when all EE are less than one, and all R are positive. If this is not the case, the modeller must check the input data again. Unlike a calibration, mass and energy balance cannot tell you if the model is a realistic description of the ecosystem, but they can assure you that results physically make sense. Moreover, the modeller can perform other checks to test the goodness of results, like checking for realistic values of P/Q or R/B, using a procedure that somehow resembles a calibration. Ecopath models are widely applied for a number of purposes, including the assessment of ecosystem trophic structure, its key groups and fishery impact, or the calculation of several indices from theoretical ecology, thermodynamics and information theory, which are very useful in evaluating ecosystem stress level. The Ecopath version used for this model is 5.1, modified with an executable file made available by Villy Christensen (Fisheries Centre, UBC) to participants in the La Paz Ecospace workshop (La Paz, Mexico, March 2007).  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  5  Modelled area and period The Northern Adriatic Sea is represented in Figure 1. The modelled area is mainly a continental shelf with sandy and muddy bottom, spanning from the Gulf of Trieste in the north southwards to the imaginary line joining the Italian city of Ancona to the Croatian island of Pago.  Figure 1. The Adriatic and Northern Adriatic Sea. The line from Italy to Croatia is the southern limit of the modelled area. The Po River plume is clearly visible in the northwest. Satellite image, modified from NASA website (www.nasa.gov).  The Eastern Croatian archipelago has steep and rocky bathymetry, while Italian coasts are gentler and sandy (Russo and Artegiani, 1996). The model does not include coastal lagoons like Venice, nor the Po River delta, since their ecological dynamics are very different from the open sea. They have consequently been considered as a reasonable approximation, external to the model. The surface of the modelled area is about 32,000 km2 and the mean depth is very shallow (29 m). The time period for collecting data and averaging is 1996–1998, but since some data had to be taken from outside this range due to limited availability, it would be more correct to speak generally of the 1990s as the reference time span. The mean water temperature for the reference period is 14.5°C and this was calculated from the Medlar/Medatlas II database (MEDAR group, 2002) by dividing the basin into layers of one meter depth and calculating a mean temperature weighted by water volumes.  6  Trophic Model of the Northern Adriatic Sea, Barausse, Duci, Mazzoldi, Artioli and Palmeri  Description of groups The model is made of 26 functional groups and 6 fleets (divided according to gear and flag). There are 2 detritus groups, 4 planktonic groups, 1 macroalgae and phanerogam group, 8 invertebrate groups, 10 fish groups and 1 seabird group. Input data before balancing and references for each group can be found in Appendix A, and balanced model data can be found in Table 3. Diet matrix, as modified after balancing, is reported in Table 1. The currency used for biomasses is t·km-2 expressed in wet weight (tWW·km-2), while the time for rates is years-1. Two detritus groups were chosen: a discard group to account for the role of discard as food in the trophic network, a detritus group to represent dead particulate organic matter in the water column and sediment (excreted and egested matter from all groups, marine snow, etc.). Biomass of detritus groups is a mandatory but not very important input, so detritus biomass was roughly calculated as the difference between particulate organic matter in the water column and the sediment, and phytoplankton and pelagic bacteria biomass. Sediment bacteria were thus not considered and assumed part of detritus. Discard biomass was estimated roughly, as it is not an important parameter. It was assumed that discard production was consumed on average in 10 days, with a linear decay (thus, it was multiplied by 10/365/2 to obtain biomass). Discard production, which is more important in the model (and in the ecosystem), was calculated from rough assumptions and unpublished data (see below). A seabird group was put in the model to account for birds feeding on discard. Planktonic groups were structured in order to simulate microbial loop, which is known to be a key pathway in energy transfers to higher trophic levels in the Northern Adriatic (Fonda Umani and Beran, 2003). Consequently, one pelagic bacteria group was included in the model. Other groups are zooplankton (micro- and meso-zooplankton, but their diet was corrected to also take into account heterotrophic nanoflagellate predation on bacteria), phytoplankton and a jellyfish group, as the Northern Adriatic has been characterised by outbreaks of several species in the last years, such as Pelagia noctiluca and Aurelia aurita (Malej, 2001). The jellyfish group biomass is quite unreliable, being based on old surveys for P. noctiluca, and so it was given a low value to avoid overestimating its predation, instead of being left out. Macroalgae and phanerogams were also included in a group, but given the scarce information on the actual bottom surface they cover, a biomass was not given and an EE value of 0.100 was chosen, following Christensen et al. (2005, p. 56). Invertebrate groups were chosen following taxonomical criteria and data availability. The filter feeding invertebrate group includes benthic organisms such as poriferans, briozoans, cnidarians, ascidians and sipunculids. Crustacea 1 are macro-crustaceans, such as mantis shrimps and Norway lobsters. Crustacea 2 consist of amphipods, isopods and benthic copepods. Biomass and production for invertebrates were mainly taken from Moodley et al. (1998) and Pranovi and Giani (1997), using data taken near the Italian coast and Po River delta, and consequently are not fully representative of the system. Diets were taken from qualitative indications found in literature (e.g. Baccetti et al., 1991) and then corrected with considerations based on personal knowledge and on personal communications by Folco Giomi (Dept. of Biology, University of Padova) and Stefano Cannicci (Dept. of Animal Biology and Genetics “Leo Pardi”, University of Firenze). The first step in constructing fish groups was to list all Northern Adriatic species, based on Riedl (1991) and personal knowledge. Then, fish species were divided into groups obtained through a cluster analysis on diet composition based on the Bray Curtis index, except for flatfish, ray and shark groups, which were defined taxonomically. After the analysis, some species were moved from one group to another, in order to obtain ecologically or commercially significant groups (e.g., fishes were moved following criteria about habitats). Grouping of species is reported in Appendix A. Diets and other fish data, such as the von Bertalanffy parameter K, used for example to calculate production and consumption using empirical equations (see references in Appendix A), were taken from existing literature. When it was not possible to find recent or good data for the Northern Adriatic Sea, literature references for the same species are based on similar ecosystems or (in some few cases) references for similar fish species.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  7  Table 1. Diet matrix of balanced model of the Northern Adriatic. Prey \ Predator  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  24  1 Seabirds 2 Sharks  0.00009 0.00002 0.00009  3 Rays  0.0001  4 European hake  0.00061 0.00015  0.000011  0.00001 0.002  0.00051  5 Zoobenthivorous fish 1  0.05  0.004  0.002  0.017  0.00001 0.002  0.006  0.015  0.0003  0.016  0.003  6 Zoobenthivorous fish 2  0.05  0.004  0.002  0.018  0.001  0.002  0.015  0.0003  0.007  0.003  7 Pelagic piscivorous fish  0.01  0.023  0.01  0.097  0.00001 0.023  0.019  0.097  0.002  0.032  0.759  0.0009  0.765  0.014  0.14  0.004  0.000099 0.00081  8 Zooplanktivorous fish  0.665 0.178  0.082  9 Omnivorous fish  0.01  0.001  0.00048 0.004  10 Benthic piscivorous fish  0.01  0.00017 0.00006 0.00031  0.000032  11 Flatfishes  0.02  0.006  0.003  0.004  0.00076  12 Cephalopods  0.005 0.204  0.023  0.003  0.008  0.539  0.508  0.642  0.024  0.0004  0.24  0.03  0.05  0.00054 0.224  13 Crustacea 1 14 Crustacea 2  0.025  0.08  0.366  15 Gastropods  0.156  0.15  0.574  0.001  0.002  0.009  0.002  0.004  0.016  0.00001  0.008  0.279  17 Filter feeding inv.  0.023  0.003  0.00013 0.00038  0.00004  21 Zooplankton  0.008  22 Macroalgae and phan.  0.087  0.027  0.321  0.107  0.002  0.000011  0.005  0.004  0.00017 0.0003  0.01  0.02  0.246  0.106  0.229  0.04  0.07  0.078  0.032  1E-04 0.068  0.179  0.139  0.032  0.068  0.047  0.009  0.00013  20 Jellyfish  0.024  0.123  0.009  19 Polychaetes  0.00041  0.017  16 Bivalves  18 Echinoderms  0.00025  0.00075  0.023  0.181  0.00075  0.045  0.002  0.467  0.346  0.009  0.037  0.99  0.008  0.003  0.2  0.00022 0.053  0.103  0.021  0.12 0.01  Sum  1  0.033  0.002  0.047  1  1  1  1  1  0.03  0.06 0.002 0.099 0.033  0.009  0.397  1  1  1  1  0.75  0.02  0.05  0.6  0.012  0.121  0.108 0.22 0.007 0.14  0.12  0.193  0.108 0.22 0.491 0.14  0.101  0.000099 0.00001 0.00071 0.0002 0.001  26 Detritus 0.06  0.002  0.106  24 Bacteria  Import  0.008  0.021  0.05  23 Phytoplankton  25 Discard  0.002  0.032  0.036 0.0004  0.04  0.15  0.19  0.00098 0.0004  0.01  0.01  0.427  0.598  0.621 0.5  0.497 0.472  0.728  0.15  0.19 0.85  1  1  1  1  1  1  1  1  1  1  1  1  8  Trophic Model of the Northern Adriatic Sea, Barausse, Duci, Mazzoldi, Artioli and Palmeri  In some extreme cases, especially for diets, data were guessed based on personal knowledge. For all piscivorous fish species, if the preyed fish species were not indicated in the reference for diet, they were chosen based on personal knowledge or assuming that all fishes with mean length smaller than the predator were preyed on in proportion to their own biomass. Diets of single species were weighted by their absolute consumptions to obtain the group diet, while production and consumption of single species were weighted by their biomasses. Consumption and production were usually taken from empirical equations (sometimes rough), from existing literature (e.g. P/B=Z=M+F) or roughly assuming GE value (usually in the middle of the range 0.1–0.3). Biomasses for fish species were taken from field surveys and, if this was not possible and if enough data were available, they were computed from Z=M+F with F=Y/B or through the proportion Y1/Y2=B1/B2, where 1 and 2 are species with similar catchability and one of the biomasses and both catches are known. These two methods are indeed very rough and can lead to estimates one order of magnitude wrong. So, if EE values computed by Ecopath for fish groups with so-estimated biomasses were low (e.g. for flatfishes during balancing), that was taken as a probable warning that computed biomasses were too high. For these reasons, in some cases it was preferred to assume an EE value instead of entering an uncertain biomass. This method can also lead to inaccurate results, especially if production is not accurately estimated, as is probably the case for zoobenthivorous fish groups. The point to keep in mind is that some biomasses are uncertain for fish groups, and this is a weak side of the model. Catch (Table 2) was taken from cross-estimates derived from unpublished data from the Chioggia fish market (Chioggia is the main fishing harbour in the modelled area), from ISTAT (Italian Institute of Statistics, www.istat.it/agricoltura/datiagri/pesca) database for Veneto, Friuli Venezia Giulia and Emilia Romagna (Italian regions) and from FAO Fisheries Statistics programme–Regional Capture Production database (www.fao.org/fi/statist/statist.asp) for Slovenia and Croatia. Unreported catch was guessed to be 40% of landings, and discard was roughly assumed to be 10% of total catch or roughly estimated from unpublished data (Dept. of Biology, University of Padova). Discard production and impact in the Northern Adriatic must surely be analysed more deeply, and our rough input data are probably underestimated (see Pranovi et al., 2001). Partitioning of Italian catch among different model gears was made according to Osservatorio Socio Economico della Pesca nell’Alto Adriatico database (www.adrifish.org) and only a percentage of about 1/3 of Croatian catch was assumed to be in the Northern Adriatic Sea.  Model balancing Model balancing was achieved by modifying input data such as to achieve mass (EE<1) and energy (R>0) conservation. Beside these physical constraints, additional checks were performed on the balanced model. Gross efficiency GE (=P/Q) values were checked to be physiologically realistic, as they usually are found in the range 0.1 and 0.3 (Christensen et al., 2005, p. 49) and EE for groups were checked to have reasonable values (EE<0.7 for phytoplankton (Opitz, 1993), relatively low values for not fully preyed or exploited groups, high values for heavily fished and preyed groups). Also, R/B values were checked to be physiologically acceptable, e.g. R/B ratio for copepods should be 50–100 year-1, as stated in Christensen et al. (2005, p. 51), and higher values are to be found for small organisms. Data that were supposed to hold higher uncertainty were modified first using a ‘search and try’ procedure until acceptable results were reached. How the input data were collected was taken into consideration when modifying them. For example, data that were known to be uncertain and underestimated were tentatively increased (e.g. cephalopod biomass does not account for Croatian and Slovenian waters) and overestimated data were lowered. Particularly, biomasses of pelagic fish groups had to be lowered, but this is in accordance with the uncertainty connected with measurement method (acoustic surveys) and with the results of Coll et al. (in press). Flatfishes showed very low EE and consequently their biomass was lowered (as they are known to be quite exploited, as can be seen also in Coll et al., in press). Some plankton data, which can be unreliable, had to be varied. In particular, zooplankton production was lowered because it appeared quite unexploited and it was known to be overestimated, and the results gave a better R/B ratio. Ecotrophic efficiency was entered for cephalopods, since production seemed too low, and also biomass increased, as previously stated. EE for macroalgae and phanerogams was raised to 0.2 because otherwise resulting biomass would have been unrealistically high. After some trials, a balanced and coherent model appeared. Final data are reported in Tables 1 and 3.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  9  Table 2. Landings and discards by fleet (t·km-2·year-1). There are 5 Italian fleets and a Slovenian-Croatian fleet. Italian fleets are hydraulic dredge, mid-water trawling (volante), bottom trawling (coccia), beam trawling (rapido), other fisheries. (Phytoplankton, bacteria, discards and detritus are here omitted.) Landings (t·km-2·year-1) Beam Bottom Mid-water CRO-SLO Hydraulic Other Group Name trawling trawling trawling Total fleet dredge fisheries (rapido) (coccia) (volante) Sea-birds 0 Sharks 0.000758 0.0028 0.000017 0.000417 0.000214 0.004 Rays 0.000517 0.00095 0.000436 0.000653 0.000135 0.003 European hake 0.0124 0.000097 0.0373 0.00108 0.051 Zoobenthivorous fish 1 0.0184 0.0213 0.00639 0.0423 0.000994 0.089 Zoobenthivorous fish 2 0.00207 0.0107 0.00428 0.038 0.000535 0.056 Pelagic piscivorous fish 0.0143 0.00978 0.0282 0.0392 0.091 Zooplanktivorous fish 0.38 0.00956 0.0701 1.513 1.973 Omnivorous fish 0.0021 0.0395 0.00719 0.0252 0.074 Benthic piscivorous fish 0.00147 0.00032 0.000222 0.0026 0.000032 0.005 Flatfishes 0.00323 0.0032 0.0416 0.0192 0.067 Cephalopods 0.0135 0.000499 0.0889 0.0166 0.0582 0.00199 0.18 Crustacea 1 0.00482 0.0197 0.0233 0.103 0.00107 0.152 Crustacea 2 0 Gastropods 0.000198 0.00686 0.123 0.00686 0.137 Bivalves 0.000508 0.0581 0.0895 0.0224 0.171 Filter feeding invertebrates 0.0000833 0 0 Echinoderms 0.0000252 0.000039 0.000010 Polychaetes 0 Jellyfish 0 Zooplankton 0 Macroalgae and phanerogams 0 Sum 0.454 0.065 0.33 0.189 0.43 1.583 3.051 Discards (t·km-2·year-1) Group Name Sea-birds Sharks Rays European hake Zoobenthivorous fish 1 Zoobenthivorous fish 2 Pelagic piscivorous fish Zooplanktivorous fish Omnivorous fish Benthic piscivorous fish Flatfishes Cephalopods Crustacea 1 Crustacea 2 Gastropods Bivalves Filter feeding invertebrates Echinoderms Polychaetes Jellyfish Zooplankton Macroalgae and phanerogams Sum  Beam trawling (rapido)  Bottom trawling (coccia)  Mid-water trawling (volante)  0.0000172 0.000052 0.000086 0.0000109 0.000033 0.000055 0.00115 0.000355 0.00248 0.00142 0.000268 0.00187 0.00107 0.00231 0.0478 0.00288 0.000719 0.000063 0.000095 0.00192 0.00064 0.00192 0.0000499 0.00889 0.00166 0.00147 0.00294 0.00734  0.000137 0.000087 0.00077 0.00248 0.00187 0.00154 0.0319 0.00216 0.000127 0.00192 0.00582 0.00294  0.000052 0.000033 0.00192 0.000355 0.000268 0.00386 0.0797 0.00144 0.000032  0.0000198 0.0000508 0.0000391 0.0000391 0.0000969  0.00457 0.00567 0.00872 0.00581 0.0127  0.00127  0.00457 0.00567 0.0134 0.00896 0.0102  0.00457 0.00567 0.00335 0.00224 0.00127  0.0000313 0.046  0.017 0.059  0.0017 0.074  0.0136 0.071  0.0017 0.071  CRO-SLO fleet  Hydraulic Other dredge fisheries  0.0000758 0.0000517 0.00124 0.00184 0.000207 0.00143 0.038 0.00021 0.000147 0.000323 0.00135 0.000482  0.000199  0.088  Total 0 0 0 0.005 0.009 0.006 0.009 0.197 0.007 0 0.007 0.018 0.015 0 0.014 0.017 0.026 0.017 0.026 0 0 0.034 0.409  10  Trophic Model of the Northern Adriatic Sea, Barausse, Duci, Mazzoldi, Artioli and Palmeri  RESULTS Trophic levels (Table 3) are in good agreement with literature (Stergiou and Karpouzi, 2002) and the highest values are found for European hake and benthic piscivorous fishes. Omnivory index (see Christensen et al., 2005) is particularly high for macrocrustaceans (Crustacea 1), which are one of the key groups in the system, as can be seen from the mixed trophic impact analysis (Ulanowicz and Puccia, 1990; see Figure 2), because of their high biomass and consumption and their wide predation spectra. The mixed trophic impact shows also that other key groups are detritus, zooplankton, phytoplankton and zooplanktivorous fish, and that the remaining fish groups have little or no impact on the network. Consequently, the ecosystem appears to have a bottom-up or more probably a wasp-waist controlled structure, possibly due to natural and anthropogenic eutrophication and the exploitation of higher trophic levels.  Figure 2. Mixed trophic impact evaluation. Fish groups have little impact on the ecosystem, which seems to be wasp-waist controlled.  The fishing fleets appear to have relatively little impact on the ecosystem, as it was found in Zucchetta et al. (2003), which argued that the Northern Adriatic Sea is in a ‘fished state’. This means that the system was intensely fished for so long that now it is in a depressed state, and it will not react to a reduction of fishing effort. This might be the case, but the real reason is probably the too high degree of aggregation in this model, which can overshadow and ‘buffer’ the intense fishing on the few commercial species. Simple system statistics calculated by Ecopath are reported in Table 4 for the balanced model. Explanations of the calculated indexes can be found in Christensen et al. (2005). Noticeably, thermodynamics indexes like total primary production–total respiration ratio are strongly influenced by bacteria presence, and so are not useful in evaluating ecosystem condition; in any case, the network shows some signs of exploitation. The structure is linear (low system omnivory index) with little recycling (low  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  11  Table 3. Balanced model data and some calculated indexes. Data in italics were calculated by Ecopath (not entered). Prod./ Cons./ Net Trophic Biomass Ecotrophic Production / Omnivory biom. Group name biom. efficiency -2 (t·km ) level efficiency consumption index -1 -1 (year ) (year ) index Seabirds 3.97 0.0088 4.61 69.34 0 0.066 0.076 0.598 Sharks 3.99 0.0262 0.289 2.877 0.619 0.1 0.118 0.177 0.382 0.2 0.235 0.11 Rays 3.62 0.0127 0.612 3.058 European hake 4.24 0.05 1.157 4.4 0.979 0.263 0.309 0.069 0.759 0.776 5.137 0.95 0.151 0.216 0.049 Zoobenthivorous fish 1 3.31 0.731 0.648 3.96 0.95 0.164 0.234 0.144 Zoobenthivorous fish 2 3.7 0.235 0.145 0.171 0.248 Pelagic piscivorous fish 3.85 2.5 0.899 6.2 0.988 0.101 0.119 0.005 Zooplanktivorous fish 3.21 14 0.89 8.776 Omnivorous fish 2.48 0.08 1.6 13.193 0.965 0.121 0.187 0.546 0.966 0.158 0.186 0.059 Benthic piscivorous fish 4.29 0.023 0.521 3.295 0.949 0.2 0.286 0.049 Flatfishes 3.25 0.150 0.888 4.439 3.936 9 0.95 0.437 0.533 0.227 Cephalopods 3.62 0.1 Crustacea 1 2.68 5.47 2.894 17.785 0.88 0.163 0.325 0.491 0.929 0.164 0.226 0.24 Crustacea 2 2.3 0.95 8.4 51.181 7.405 1.699 9.51 0.95 0.179 0.447 0.253 Gastropods 2.28 0.521 0.223 0.637 0.223 Bivalves 2.29 25.599 1.415 6.35 0.924 0.2 0.364 0.251 Filter feeding inv. 2.5 7.652 0.761 3.804 Echinoderms 2.44 8.847 0.803 2.514 0.362 0.319 0.581 0.323 0.595 0.115 0.256 0.137 Polychaetes 2.15 26.989 1.644 14.27 0.151 0.333 0.417 0.285 Jellyfish 3.01 1.02 8.43 25.3 0.95 0.344 0.454 0.174 Zooplankton 2.21 3.279 55 160 Macroalgae and phan. 1 38.198 1.699 0.2 0 0.38 0 Phytoplankton 1 12.76 169.28 0.409 0.19 0.237 0 Bacteria 2 4.014 127.241 670 0.974 0 Discard 1 0.006 0.999 0.354 Detritus 1 361.93 -  predatory cycling index), which implies a ‘developmental’ stage sensu Odum (1969). Gross efficiency of fishery is relatively high, even if the system is eutrophicated. The mean trophic level is low and comparable to existing literature values (Coll et al., in press), and is reflective of the fact that the main landings in the area are small pelagics and invertebrates. Total primary production–total biomass ratio is high and shows that the ecosystem is immature and strongly productive, because of eutrophication, even if energy fluxes are mainly (65%) based on detritus, which is usually seen as a sign of maturity. However, the causes are probably to be found in the shallowness of the basin and in the variable hydrodynamic regime, which increase sediment-water column interactions. Energy transfer efficiency TE between trophic levels is 12.6% on average (calculated as geometric mean of trophic levels from II to IV), and the primary producer–based transfer is more efficient (TE=13%) than the detritus-based one (TE=12.5%).  DISCUSSION The Northern Adriatic Sea appears to be a quite immature ecosystem. However, it is not clear in which degree its stressed structure is due to anthropogenic causes (eutrophication, fishing) or to its natural characteristics (shallowness, low residence time, Po River input). Consequently it is difficult to quantify ecosystem health. This model provides some interesting insights on trophic structure, key functional groups and energy flows in the system, but it would probably benefit from a lower degree of aggregation and from more precise fish biomass values. Time simulation could be surely useful in order to test hypotheses on ecosystem control and on the importance of fisheries.  12  Trophic Model of the Northern Adriatic Sea, Barausse, Duci, Mazzoldi, Artioli and Palmeri  Table 4. System statistics for the tropic model of the Northern Adriatic Sea. Attribute Sum of all consumption Sum of all exports Sum of all respiratory flows Sum of all flows into detritus Total system throughput Sum of all production Mean trophic level of the catch Gross efficiency of fishery (catch/net p.p.) Calculated total net primary production Total primary production/total respiration Net system production Total primary production/total biomass Total biomass/total throughput Total biomass (excluding detritus) Total catches Connectance Index System Omnivory Index Throughput cycled (excluding detritus) Predatory cycling index Throughput cycled (including detritus) Finn's cycling index Finn's mean path length Finn's straight-through path length Finn's straight-through path length  Value 4203.986 6.561 2218.795 2894.021 9324 3070 3.11 0.001555 2224.912 1.003 6.117 13.852 0.017 160.623 3.46 0.372 0.209 38.62 1.09 2138.66 22.94 4.19 1.573 3.229  Units t·km-2·year-1 t·km-2·year-1 t·km-2·year-1 t·km-2·year-1 t·km-2·year-1 t·km-2·year-1  t·km-2·year-1 t·km-2·year-1 year-1 year t·km-2 t·km-2·year-1  t·km-2·year-1 % of throughput w/o detritus t·km-2·year-1 % of total throughput without detritus with detritus  ACKNOWLEDGEMENTS This work was developed under the EU-sponsored INCOFISH project (contract INCO-003739). We thank Prof. Corrado Piccinetti (BES, University of Bologna) for cephalopod MEDITS data, and Folco Giomi (Dept. of Biology, University of Padova) and Stefano Cannicci (Dept. of Animal Biology and Genetics “Leo Pardi”, University of Firenze) for their advice.  REFERENCES Abdel Aziz, S.H. 1992. The use of vertebral rings of the brown ray Raja miraletus (Linnaeus, 1758) off Egyptian mediterranean coast for estimation of age and growth. Cybium, 16 (2): 121-132. Arreguín-Sánchez, F., Seijo, J.C. and Valero-Pacheco, E. 1993. An application of ECOPATH II to the north continental shelf ecosystem of Yucatan, Mexico. p. 269-278. In V. Christensen and D. Pauly (eds.) Trophic models of aquatic ecosystems. ICLARM Conf. Proc. 26, 390 pp. Artuz, M.L. 2005. The diet and food consumption of whiting Merlangius merlangus merlangus (Linné) 1758 in the Sea of Marmara. Hidrobiologica, 1. Azevedo, J.M.N. and Simas, A.M.V. 2000. Age and growth, reproduction and diet of a sublittoral population of therock goby Gobius paganellus (Teleostei, Gobiidae). Hydrobiologia, 440: 129–135. Azzali, M., De Felice, A., Luna, M., Cosimi, G. and Parmigiani, F. 2002. The state of the Adriatic Sea centered on the small pelagics fish populations. Marine Ecology, 23, Supplement 1: 78-91. Baccetti, B., Baldaccini, N.E., Bedini, C., Brandymar, P., Capanna, E., Chieffi, G., Coboldi, M., Ferraguti, M., Ghirardelli, E., Ghiretti, F., Giusti, F., Gigolo, A., Mainardi, D., Minelli, A., Papi, F., Parriniello, N., Ricci, N., Ruffo, S., Sarà, M., Scali, V. and Zullini, A. 1991. Zoologia 2 – Trattato italiano, Editoriale Grasso, Bologna. Bartulovic, V., Glamuzina, B., Conides, A., Dulcic, J., Lucic, D., Njire, J. and Kozul, V. 2004. Age, growth, mortality and sex ratio of sand smelt, Atherina boyeri Risso, 1810 (Pisces: Atherinidae) in the estuary of the Mala Neretva River (middle-eastern Adriatic, Croatia). Journal of Applied Ichthyology, 20: 427-430. Benovic, A. 2000. Zooplankton biomass and fish production in the Adriatic. CIESM workshop series, Kerkyra (Greece), 26th august30th July 2000. Brey, T. 1999. Growth performance and mortality in aquatic macrobenthic invertebrates. Advances in Marine Biology, 35: 153-223.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  13  Brey, T. 2001. Population dynamics in benthic invertebrates. A virtual handbook. Version 01.2. http://www.awibremerhaven.de/Benthic/Ecosystem/FoodWeb/Handbook/main.html. Alfred Wegener Institute for Polar and Marine Research, Germany. Cabrini, M., Andri, M., Cok, S., Pecchiar, I. and Fonda Umani, S. 2002. Carbon partitioning among the first trophic levels in the North Western Adriatic basin. Chemistry and Ecology 18(1-2): 95-105. Christensen, V. and Walters, C. 2004. Ecopath with Ecosim: methods, capabilities and limitations. Ecological Modelling, 172(2-4): 109-139. Christensen, V., Walters, C.J. and Pauly, D. 2005. Ecopath with Ecosim: a user’s guide. Fisheries Centre, University of British Columbia, Vancouver. November 2005 edition, 154 pp. (available online at www.ecopath.org). Cingolani, N., Santojanni, A., Arneri, E., Berlardinelli, A., Colella, S., Donato, F., Giannetti, G., Sinovcic, G. and Zorica, B. 2004a. Sardine (Sardina pilchardus, Walb.) stock assessment in the Adriatic Sea: 1975-2003. Paper presented at the GFCM-SAC Working Group on Small Pelagic Species, GCP/RER/010/ITA/OP-13. Cingolani, N., Santojanni, A., Arneri, E., Berlardinelli, A., Colella, S., Donato, F., Giannetti, G., Sinovcic, G. and Zorica, B. 2004b. Anchovy (Engraulis encrasicolus, L.) stock assessment in the Adriatic Sea: 1975-2003. Paper presented at the GFCM-SAC Working Group on Small Pelagic Species, GCP/RER/010/ITA/OP-14. Coelho, M., Domingues, P., Balguerias, E., Femandez, M. and Andrade, J.P. 1997. A comparative study of the diet of Loligo vulgaris (Lamarck, 1799) (Mollusca: Cephalopoda) from the south coast of Portugal and the Saharan Bank (Central-East Atlantic). Fisheries Research, 29: 245-255. Coll, M., Santojanni, A., Palomera, I., Tudela, S. and Arneri, E. In press. An ecological model of the Northern and Central Adriatic Sea: analysis of ecosystem structure and fishing impacts. Cortés, E. 1999. Standardized diet compositions and trophic levels of sharks. ICES Journal of Marine Science, 56: 707-717. Curtis, J.M.R. and Vincent, A.C.J. 2006. Life history of an unusual marine fish: survival, growth and movement patterns of Hippocampus guttulatus (Cuvier 1829). Journal of Fish Biology, 68: 707–733. Danovaro, R. 2003. Interazioni DOM/POM nei processi di formazione dei macroaggregati mucillaginosi in Adriatico Settentrionale: ruolo dei virus e dei batteri. Istituto per la ricerca scientifica e tecnologica applicata al mare, Programma di monitoraggio e studio sui “Processi di formazione delle Mucillaggini nell’Adriatico e nel Tirreno” – MAT, Rapporto finale giugno 1999-maggio 2003. Giugno 2003. CD-ROM. Danovaro, R., Manini, E. and Fabiano, M. 2001. Exoenzymatic activity and organic matter composition in sediments of the Northern Adriatic sea: response to a river plume. Microbial Ecology, 44: 235-251. Degobbis, D., Vojvodic, V., Ahel, M., Tepic, N., Cosovic, B., Muhin, N., Zovko, Z., Zangrando, V., Giani, M. and Savelli, F. 2003. Ruolo del carbonio organico, carboidrati e azoto particellato negli eventi di mucillagini lungo il transetto A nel periodo 1999-2002. Istituto per la ricerca scientifica e tecnologica applicata al mare, Programma di monitoraggio e studio sui “Processi di formazione delle Mucillaggini nell’Adriatico e nel Tirreno” – MAT, Rapporto finale giugno 1999-maggio 2003. Giugno 2003. CD-ROM. Duarte, C.M. and Chiscano, C.L. 1999. Seagrass biomass and production: a reassessment. Aquatic Botany, 65: 159-174. Dulcic, J. and Glamuzina, B. 2006. Length–weight relationships for selected fish species from three eastern Adriatic estuarine systems (Croatia). Journal of Applied Ichthyology, 22: 254– 256. Dulcic, J. and Kraljevic, M. 1995. Age, growth and mortality of damselfish (Chromis chromis L.) in the eastern middle Adriatic. Fisheries Research, 22: 255-264. Dulcic, J. and Kraljevic, M. 1997. Corrigendum to “Weight-length relationships for 40 fish species in the eastern Adriatic ( Croatian waters)” [Fisheries Research 28 ( 1996) 243- 25 l]. Fisheries Research, 30: 169-171. Dulcic, J., Kraljevic, M., Grbec, B. and Cetinic, P. 2000. Age, growth and mortality of blotched picarel Spicara maena L. (Pisces: Centracanthidae) in the eastern central Adriatic. Fisheries Research, 48: 69-78. Dulcic, J., Pallaoro, A., Cetinici, P., Kraljevic, M., Soldo, A. and Jardas, I. 2003. Age, growth and mortality of picarel, Spicara smaris L. (Pisces: Centracanthidae), from the eastern Adriatic (Croatian coast). Journal of Applied Ichthyology, 19 (1). FAO Adriamed–Loligo vulgaris, 2006. Visited November 2006: www.faoadriamed.org/html/Species/LoligoVulgaris.html. FAO Adriamed-Lophius budegassa, 2006. Visited November 2006: www.faoadriamed.org/html/Species/ LophiusBudegassa.html. FAO Adriamed-Lophius piscatorius, 2006. Visited November 2006: www.faoadriamed.org/html/Species/ LophiusPiscatorius.html. FAO Adriamed-Merlangius merlangus, 2006. Visited November 2006: www.faoadriamed.org/html/Species/Merlangius Merlangus.html. FAO Adriamed-Merluccius merluccius, 2006. Visited November 2006: www.faoadriamed.org/html/Species/Merluccius Merluccius.html. FAO Adriamed-Mullus barbatus, 2006. Visited November 2006: www.faoadriamed.org/html/Species/MullusBarbatus.html. FAO Adriamed-Pagellus erythrinus, 2006. Visited November 2006: www.faoadriamed.org/html/Species/ PagellusErythrinus.html. FAO Adriamed-Sepia officinalis, 2006. Visited November 2006: www.faoadriamed.org/html/Species/SepiaOfficinalis.html. FAO Adriamed-Solea vulgaris, 2006. Visited November 2006: www.faoadriamed.org/html/Species/SoleaVulgaris.html.  14  Trophic Model of the Northern Adriatic Sea, Barausse, Duci, Mazzoldi, Artioli and Palmeri  Filiz, H. and Bilge, G. 2004. Length–weight relationships of 24 fish species from the North Aegean Sea, Turkey. Journal of Applied Ichthyology, 20: 431–432. Filiz, H. and Mater, S. 2002. A Preliminary Study on Length-Weight Relationships for Seven Elasmobranch Species from North Aegean Sea, Turkey. E.U. Journal of Fisheries and Aquatic sciences, 19 (3-4): 401-409. Fonda Umani, S. and Beran, A. 2003. Seasonal variations in the dynamics of microbial plankton communities: first estimates from experiments in the Gulf of Trieste, Northern Adriatic Sea. Marine Ecology Progress Series, 247: 1-16. Fonda Umani, S., De Olazabal, A., Parlato, S., and Boglich Perasti, B. 2003. Evoluzione del popolamento mesozooplanctonico lungo i transetti B e C. Istituto per la ricerca scientifica e tecnologica applicata al mare, Programma di monitoraggio e studio sui “Processi di formazione delle Mucillaggini nell’Adriatico e nel Tirreno” – MAT, rapporto finale giugno 1999-maggio 2003. Giugno 2003. CD-ROM. Fonda Umani, S., Tirelli, V., Beran, A. and Guardiani, B. 2005. Relationships between microzooplankton and mesozooplankton: competition versus predation on natural assemblages of the Gulf of Trieste (northern Adriatic Sea). Journal of Plankton Research, 27 (10): 973-986. Froese, R. and Pauly, D., Editors. 2007. FishBase. World Wide Web electronic publication. www.fishbase.org, version (05/2007). Visited from June 2006 to May 2007. Giani, M., Baldan, C., Savelli, F., Zangrando, V., Mecozzi, M. and Amici, M. 2003. Distribuzione e variazioni temporali della sostanza organica disciolta e particellata nell’Adriatico sulle sezioni Senigalli-isola di Sansego e Cesenatico-Capo Promontore. Istituto per la ricerca scientifica e tecnologica applicata al mare, Programma di monitoraggio e studio sui “Processi di formazione delle Mucillaggini nell’Adriatico e nel Tirreno” – MAT, rapporto finale giugno 1999-maggio 2003. Giugno 2003. CD-ROM. Guenette, S., Morato, T. 1997. The Azores Archipelago. In: Fisheries impacts on North Atlantic ecosystems: models and analyses. S. Guénette, Christensen, V. and Pauly, D. (eds.). Fisheries Centre Research Reports 9 (4). Guidetti, P., Lorenti, M., Buia, M.C. and Mazzella, L. 2002. Temporal Dynamics and Biomass Partitioning in the Three Adriatic Seagrass Species: Posidonia oceanica, Cymodocea nodosa, Zostera marina. Marine Ecology, 23 (1): 51-67. Ismen, A. 2003. Age, growth, reproduction and food of common stingray (Dasyatis pastinaca L., 1758) in Iskenderun Bay, the eastern Mediterranean. Fisheries Research, 60: 169–176. Jukic-Peladic, S., Vrgoc, N., Krtulovic-Sifner, S., Piccinetti, C., Piccinetti-Manfrin, G., Marano, G. and Ungaro, N. 2001. Long-term changes in demersal resources of the Adriatic Sea: comparison between trawl surveys carried out in 1948 and 1998, Fisheries Research, 53: 95-104. Kallianiotis, A., Torre, M. and Argyri, A. 2005. Age, growth, mortality, reproduction and feeding habits of the striped seabream, Lithognathus mormyrus (Pisces: Sparidae), in the coastal waters of the Thracian Sea, Greece. Scientia Marina, 69 (3): 391-404. Khailov, K.M. and Burlakova, Z.P. 1969. Release of Dissolved Organic Matter by Marine Seaweeds and Distribution of Their Total Organic Production to Inshore Communities. Limnology and Oceanography, 14 (4): 521-527. Kozul, V., Skaramuca, B., Kraljevic, M., Dulcic, J. and Glamuzina, B. 2001. Age, growth and mortality of the Mediterranean amberjack Seriola dumerili (Risso 1810) from the south-eastern Adriatic Sea. Journal of Applied Ichthyology, 17: 134-141. Kraljevic, M., Dulcic, J., Cetinic, P. and Pallaoro, A. 1996. Age, growth and mortality of the striped sea bream, Lithognathus mormyrus L., in the Northern Adriatic. Fisheries Research, 28: 361-370. La Ferla, R., Azzaro, M., Zaccone, R. and Caruso, G. 2002. Microbial respiratory and ectoenzymatic activities in the Northern Adriatic sea (Mediterranean sea). Chemistry and Ecology, 18(1-2): 75-84. Labropoulou, M., Tserpes, G. and Tsimenides, N. 1998. Age, Growth and Feeding Habits of the Brown Comber Serranus hepatus (Linnaeus, 1758) on the Cretan Shelf. Estuarine, Coastal and Shelf Science, 46: 723-732. Link, J.S., Griswold, C.A., Methratta, E.T. and Gunanrd, J. (eds). 2006. Documentation for the Energy Modeling and Analysis eXercise (EMAX), Northeast Fisheries Science Centre Reference Document 06-15, August 2006. Lipej, L., Bonaca, M.O. and Sisko, M. 2003. Coastal Fish Diversity in Three Marine Protected Areas and One Unprotected Area in the Gulf of Trieste (Northern Adriatic). Marine Ecology, 24 (4): 259–273. Malej, A. 1989. Behaviour and trophic ecology of the jellyfish Pelagia noctiluca (Forsskål, 1775). Journal of Experimental Marine Biology and Ecology, 126: 259-270. Malej, A. 2001. Are irregular plankton phenomena getting more frequent in the northern Adriatic Sea?. Gelatinous zooplankton outbreaks, theory and practice, CIESM workshop series, Naples (Italy), 29th august-1st september 2001. Malej, A. and Malej, A. Jr. 2004. Invasion of the jellyfish Pelagia noctiluca in the Northern Adriatic: a non-success story. H. Dumont et al. (eds.) Aquatic invasion in the Black, Caspian and Mediterranean Seas: 273-285. MEDAR group, 2002. MEDATLATS/2002 database. Mediterranean and Black Sea database of temperature, salinity and biochemical parameters. Climatological Atlas. 4 CD-ROM, IFREMER edition. Moodley, L., Heip, C.H.R. and Middelburg, J.J. 1998. Benthic activity in sediments of the northwestern Adriatic Sea: sediment oxygen consumptiom, macro- and meiofauna dynamics. Journal of Sea Research, 40: 263-280. Munda, I.M. 1990. Resources and possibilities for exploitation of North Adriatic seaweeds. Hydrobiologia, 204-205: 309-315. Munda, I.M. 1993. Change and degradation of seaweeds stands in the Northern Adriatic seaweeds. Hydrobiologia, 260-261: 239-253.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  15  Oddo, P., Pinardi, N. and Zavatarelli, M. 2005. A numerical study of interannual variability of the Adriatic Sea (2000-2002). Science of the Total Environment, 353: 39-56. Odum, E.P. 1969. The strategy of ecosystem development. Science, 164: 262-270. Opitz, S. 1993. A quantitative model of the trophic interactions in a Caribbean coral reef ecosystem. p. 259-267. In V. Christensen and D. Pauly (eds.) Trophic models of aquatic ecosystems. ICLARM Conf. Proc. 26, 390 pp. Opitz, S. 1996. Trophic Interactions in Caribbean Coral Reef. ICLARM Tech. Rep. 43, ICLARM, Makati City, Philippines, 341 pp. Osservatorio Socio Economico della Pesca nell’Alto Adriatico database (www.adrifish.org, La pesca e l’aquacultura nel Veneto, La pesca e l’aquacultura in Emilia Romagna, La pesca e l’aquacultura in Friuli Venezia Giulia). Pallaoro, A., Cetinic, P., Dulcic, J., Jardas, I. and Kraljevic, M. 1998. Biological parameters of the saddled bream Oblada melanura in the eastern Adriatic. Fisheries Research, 38: 199-205. Pallaoro, A. and Jardas, I. 2003. Some biological parameters of the peacock wrasse Symphodus (Crenilabrus) tinca (L. 1758) (Pisces: Labridae) from the middle eastern Adriatic (Croatian coast). Scientia Marina, 67 (1): 33-41. Pinnegar, J.K. and Polunin, N.V.C. 2004. Predicting indirect effects of fishing in Mediterranean rocky littoral communities using a dynamic simulation model. Ecological Modelling, 172: 249-267. Pranovi, F. and Giani, M. 1997. Popolazioni macrobentoniche in alcuni siti dell’Adriatico settentrionale: composizioni e biomasse. Pesca e ambiente nella laguna di Venezia e nell’alto Adriatico, “Il Leggio” Libreria Editrice, Sottomarina – Chioggia (VE). Pranovi, F., Raicevich, S., Franceschini, G., Torricelli, P. and Giovanardi, O. 2001. Discard analysis and damage to non-target species in the “rapido” trawl fishery. Marine Biology, 139: 863-875. Ragonese, S., Fiorentino, F., Garofalo, G., Gristina, M., Levi, D., Gancitano, S., Giusto, G.B., Rizzo, P., Sinacori, G., n.d. Distribution, abundance and biological features of picarel (Spicara flexuosa), Mediterranean (Trachurus mediterraneus) and Atlantic (T. trachurus) horse mackerel based on experimental bottom-trawl data (MEDITS, 1994–2002) in the Strait of Sicily. MedSudMed Technical Documents, 5. Riedl, R. 1991. Flora e fauna del mediterraneo, Franco Muzzio & c. Editore, Padova. Rodriguez-Cabello, C. and Sanchez, F. 2005. Mortality estimates of Scyliorhinus canicula in the Cantabrian sea using tag-recapture data. Journal of Fish Biololgy, 66 (4) pp. 1116-1126. Russo, A. and Artegiani, A. 1996. Adriatic sea hydrography. Scientia Marina, 60(2): 33-43. Santic, M., Jardas, I. and Pallaoro, A. 2002. Age, growth and mortality rate of horse mackerel, Trachurus trachurus (L.), living in the eastern central Adriatic. Periodicum Biologorum, 104 (2): 165-173. Sartor, P., Sbrana, M., Ungaro, N., Marano, A.C., Piccinetti, C. and Manfrin, G. 2002. Distribution and abundance of Citharus linguatula (Linnaeus, 1758), Lepidorhombus boscii (Risso, 1810) and Solea vulgaris, Quensel, 1806 (Osteichthyes, Pleuronectiformes) in the Mediterranean sea. Scientia Marina, 66 (2): 83-102. SIBM, 2006. Standardizzazione delle metodiche di determinazione specifica e valutazione degli stocks delle razze Società italiana di Biologia Marina, Comitato Necton e Pesca, GRUppo Nazionale Risorse Demersali. Sinovcic, G. 2001. Small pelagic fish from the Croatian fishing grounds. In Mannini, P., Massa, F. and Milone, N. (eds), Priority topics related to small pelagics fishery resources of the Adriatic Sea, GCP/RER/010/ITA/TD-03, FAO-AdriaMed. Sinovcic, G., Franicevic, M., Zorica, B. and Cikes, V. 2004. Length–weight and length–length relationships for 10 pelagic fish species from the Adriatic Sea (Croatia). Journal of Applied Ichthyology, 20: 156–158. Spedicato, M.T., Greco, S., Sophronidis, K., Lembo, G., Giordano, D. and Argyri, A. 2002. Geographical distribution, abundance and some population characteristics of the species of the genus Pagellus (Osteichthyes: Perciformes) in different areas of the Mediterranean. Scientia Marina, 66 (2). Stergiou, K.I. and Karpouzi, V.S. 2002. Feeding habits and trophic levels of Mediterranean fish. Reviews in Fish Biology and Fisheries, 11: 217–254. Ulanowicz, R.E. and Puccia, C.J. 1990. Mixed trophic impacts in Ecosystems. Coenoses, 5: 7-16. Vannucci, S. 2005. Ecologia di alcune specie di Rajidae nel Mar Ligure meridionale con particolare riferimento all’alimentazione’, Thesis, Università degli Studi di Pisa. Visited October 2006: http://etd.adm.unipi.it/theses/available/etd-06292005122910/unrestricted/Tesi_Simona_Vannucci.pdf. Villamil, M.M., Lorenzo, J.M., Pajuelo, J.G., Ramos, A. and Coca, J. 2002. Aspects of the life history of the salema, Sarpa salpa (Pisces, Sparidae), off the Canarian Archipelago (central-east Atlantic). Environmental Biology of Fishes, 63: 183-192. Vrgoc, N., Krstulovic Sifner, S., Dadic, V. and Jukic-Peladic, S. 2006. Demographic structure and distribution of John Dory, Zeus faber L. 1758, in the Adriatic Sea. Journal of Applied Ichthyology, 22: 205–208. Zorica, B., Sinovcic, G., Pallaoro, A. and Cikes Kec, V. 2006. Reproductive biology and length–weight relationship of painted comber, Serranus scriba (Linnaeus, 1758), in the Trogir Bay area (middle-eastern Adriatic). Journal of Applied Ichthyology, 22: 260– 263. Zucchetta, M., Libralato, S., Granzotto, A., Pranovi, F., Raicevich S. and Torricelli, P. 2003. Modelling Approach for the Evaluation of the Efficacy of MPA in the Northern Adriatic Sea. Proceedings of the sixth international conference on the Mediterranean coastal environment. MEDCOAST 03. Erdal Özhan (Editor), 7th-11th October 2003, Ravenna, Italy.  16  Trophic Model of the Northern Adriatic Sea, Barausse, Duci, Mazzoldi, Artioli and Palmeri  Appendix A. Input data for Northern Adriatic Sea Ecopath model (before balancing). Detritus  Value  B  361.930 t·km-2  Discard  Value  B  0.0056 t·km-2  Bacteria  Value  Source  B  4.014 t·km-2  La Ferla et al. (2002); Danovaro (2003)  P/B  127.241 year-1  La Ferla et al. (2002); Danovaro (2003)  Q/B  760.018 year-1  References  Notes Calculated as the difference Danovaro et al. between particulate organic (2001); Degobbis matter (in water and 1 cm 10 gWW/gC (Opitz, 1996) et al. (2003); sediment) and biomass of Giani et al. (2003) phytoplankton and pelagic bacteria Source Conversion factors Notes Consumed in 10 days, linear decaying  La Ferla et al. (2002); Danovaro (2003)  GS  0.200  Link et al. (2006)  DC  Table 1  Link et al. (2006)  Phytoplankton  Value  References  B  12.760 t·km-2  MEDAR group (2002)  P/B  169.280 year-1  MEDAR group (2002)  Macroalgae and Value phanerogams  References  0.100 t·km-2  Christensen et al. (2005)  P/B  1.699 year-1  Khailov and Burlakova (1969); Guidetti et al. (2002); Duarte and Chiscano (1999); Munda (1990); Munda (1993)  Zooplankton  Value  B  3.279 t·km-2  P/B  90.557 year-1  EE  Conversion factors  Source Cabrini et al. (2002); Fonda Umani et al. (2003) Benovic (2000); Pinnegar and Polunin (2004)  Conversion factors  Notes  10 gWW/gC (Link et al., 2006)  Computed from equation 2 using P/B and R/B from La Ferla et al. (2003), and GS from Link et al. (2006), and then re-calculated as (P/B)/(GE) from input P/B and GE from La Ferla et al. (2003)  Conversion factors  Notes  10 gWW/gC, 400 gWW/gChl-a (Link et al., 2006)  Averaged over 1990s Averaged over 1990s  Conversion factors  Notes No reliable biomass estimates  Considered biomass above ground. Phanerogams: Z. marina, (Arreguín-Sánchez et al., P. oceanica, C. nodosa. P/B s of 1993); 4.4 gWW/gDW for macroalgae and phanerogams weighted on biomasses and Zostera marina, 5 gWW/gDW for Cymodocea assuming a covered bottom surface ratio of about 1.87, nodosa (Brey, 2001) following bathymetric considerations on vegetated bottom Conversion factors Notes Computed as sum of Mesozooplankton: 5.6 gWW/gDW; 2,222 gDW/gC microzooplankton (1996-1998) and mesozooplantkon (in 1999(Brey, 2001); Microzooplankton: 5.556 2000) gWW/gDW; 2.174 gDW/gC Computed as mean weighted on (Link et al., 2006) biomasses of microzooplankton and mesozooplankton P/B's 7.7 gWW/gDW for  Posidonia oceanica  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  Computed choosing GE=0.5 for microzooplankton and GE=0.3 for mesozooplankton (Pinnegar and Polunin, 2004).  -1  Q/B  197.759 year  GS  0.243  Computed as mean weighted on biomasses of microzooplankton and mesozooplankton GS's  Link et al. (2006) Fonda Umani and Beran (2003); Fonda Umani et al. (2005); Link et al. (2006) Source Conversion factors  DC  Table 1  Jellyfish  Value  B  1.020 t·km-2  P/B  8.430 year-1  Malej (1989)  Q/B  25.300 year-1  Malej (1989)  GS  0.200  DC  Table 1  Malej (1989) Coll et al. (in press)  Polychaetes  Value  Source  B  26.989 t·km-2  Moodley et al. (1998)  P/B  1.644 year-1  Moodley et al. (1998)  -1  17  0.0049 gC/gWW (Malej, 1989)  Conversion factors  Arreguín-Sánchez et al. (1993); 15.2124 gWW/gC (Brey, Pinnegar and 2001) Polunin (2004) Coll et al. (in press); Link et al. (2006) Baccetti et al. (1991) Source Conversion factors  14.270 year  GS  0.550  DC  Table 1  Echinoderms  Value  B  8.847 t·km-2  Moodley et al. (1998)  P/B  0.803 year-1  Moodley et al. (1998)  Q/B  -1  2.514 year  GS  0.450  DC  Table 1  Filter feeding invertebrates  Value  B  7.652 t·km-2  P/B  0.761 year-1  Pinnegar and Polunin (2004)  Notes 1984 (a low value was taken, since P. noctiluca in 1990s is lower than 1980s)  Malej and Malej (2004)  Q/B  Detritus percentage in diet taken from Link et al. (2006). Value similar to the one in Coll et al. (in press)  Notes  Average of values corrected to account for temperature difference, with empirical equation from Opitz (1996) Mean value  Notes  Value corrected to account for temperature difference, with 26.7 gWW/gC (Brey, 2001) empirical equation from Opitz (1996)  Coll et al. (in press); Link et al. (2006) Baccetti et al. (1991) Source Moodley et al. (1998) Moodley et al. (1998)  Conversion factors 0.043 gC/gWW (Moodley  et al., 1998)  Notes Calculated as ‘rest’  18  Trophic Model of the Northern Adriatic Sea, Barausse, Duci, Mazzoldi, Artioli and Palmeri  Q/B  3.804 year-1  GS  0.450  DC  Table 1  Bivalves  Value  Assuming GE=0.2 Coll et al. (in press); Link et al. (2006) Baccetti et al. (1991) Source Conversion factors  B  25.599 t·km-2  P/B  1.415 year-1  Q/B  -1  6.350 year  GS  0.650  DC  Table 1  Gastropods  Value  EE  0.950  P/B  1.699 year-1  Q/B  -1  9.510 year  GS  0.600  DC  Table 1  Crustacea 1  Value  Pranovi and Giani (1997) Moodley et al. (1998) Opitz (1996)  Weighted station depths according to Northern Adriatic bathymetry Based on Corbula gibba 0.435 gWW/gWW+shell (Brey 2001)  Notes  Pinnegar and Polunin (2004)  Value corrected to account for temperature difference, with empirical equation from Opitz (1996)  B  5.384 t·km-2  Pranovi and Giani (1997)  P/B  2.894 year-1  Moodley et al. (1998)  Q/B  17.785 year  GS  0.500  DC  Table 1  Crustacea 2  Value  B  1.010 t·km-2  P/B  7.908 year-1  Value corrected to account for temperature difference, with empirical equation from Opitz (1996)  Coll et al. (in press); Link et al. (2006) Baccetti et al. (1991) Source Conversion factors Christensen et al. (2005) Opitz (1996); Pinnegar and Polunin (2004)  Link et al. (2006) Baccetti et al. (1991) Source Conversion factors  -1  Notes  No reliable biomass estimates Assuming GE=0.178, as in the references  Notes Computed as difference of crustacean biomass (in reference) and Crustacea 2 biomass (in model). In reference, station depths were weighted according to Northern Adriatic bathymetry.  10.215 gWW/gC (Brey, 2001)  Pinnegar and Polunin (2004) Link et al. (2006) Baccetti et al. (1991) Source Conversion factors  Computed subtracting Crustacea 2 production (from model) to total crustacean production in reference Value corrected to account for temperature difference, with empirical equation from Opitz (1996)  Notes  Coll et al. (in press) Assumed GE=0.154 as in Coll et al. (in press)  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  Q/B  51.181 year  Coll et al. (in press)  GS  0.275  Coll et al. (in press)  DC  Table 1  Cephalopods  Value  Source  0.030 t·km-2  Unpublished MEDITS trawl surveys data (1996-1998) from prof. C. Piccinetti (BES, University of Bologna)  B  -1  Value corrected to account for temperature difference, with empirical equation from Opitz (1996)  Conversion factors  FAO Adriamed -  Loligo vulgaris  P/B  2.100 year  -1  (2006); FAO Adriamed - Sepia Officinalis (2006); Riedl (1991) 0.2 gDW/gWW; 22.03 kJ/gDW (Brey, 2001)  Q/B  8.470 year  Guenette and Morato (1997)  GS  0.180  Coll et al. (in press); Link et al. (2006)  DC  Table 1  Baccetti et al. (1991); Coehlo et al. (1997); FAO Adriamed - Sepia officinalis (2006); Riedl (1991)  Flatfishes  Value  Source  -1  19  Notes  Weighted mean for production of Loligo vulgaris (Z=M+F, M from eq. B2 in Brey, 1999), Eledone moschata and Sepia officinalis (empirical equation from Hoenig (1983, cit. in Christensen et al., 2005, p. 39) Value for Loligo forbesi corrected to account for temperature difference, with empirical equation from Opitz (1996)  Notes Species: Psetta maxima, Platichthys flesus, Lepidorhombus  boschi, Microchirus variegatus, Pegusa impar, Synapturichthys  -2  B  0.404 t·km  P/B  0.888 year-1  Q/B  4.439 year-1  Froese and Pauly kleinii kleini, Solea solea, Lepidorhombus whiffiagonis, (2007); Sartor et Citharus linguatula. MEDITS trawl surveys. Proportion al. (2002) between catches and biomasses; Z=M+Y/B (see Description of groups) Coll et al. (in press) Dulcic and Glamuzina (2006); Fao Adriamed -  Solea Vulgaris  (2006); Froese and Pauly (2007)  GS  0.300  Coll et al. (in press); Link et al. (2006)  DC  Table 1  Froese and Pauly (2007)  Assumed GE=0.2  Empirical eq. 17 in Christensen et al. (2005)  20  Benthic piscivorous fishes  B  Trophic Model of the Northern Adriatic Sea, Barausse, Duci, Mazzoldi, Artioli and Palmeri  Value  Source  0.021 t·km-2  FAO Adriamed Lophius Piscatorius (2006); Species: Belone belone, Lichia amia, Seriola dumerilii, FAO Adriamed - Trachinotus ovatus, Conger conger, Lophius piscatorius, Sarda sarda, Scorpaena scrofa, Serranus cabrilla, Dentex dentex, Lophius Trachinus draco, Trachinus Aranaeus, Uranoscopus scaber, Budegassa Zeus faber, Lophius Budegassa. MEDITS trawl surveys. (2006); Froese and Pauly (2007); Proportion between catches and biomasses; Z=M+Y/B (see Jukic-Peladic et al. description of groups) (2001); Vrgoc et al. (2006)  -1  P/B  0.521 year  Q/B  3.295 year-1  GS  0.150  DC  Table 1  Omnivorous fishes  Value  B  Notes  FAO Adriamed - Z=M+F; M from eq. B3 in Brey (1999). Assumed GE=0.2 for Lophius C. conger, S. sarda, S. scrofa, D. dentex; Inverted empirical Piscatorius (2006); eq. 19 in Christensen et al. (2005) FAO Adriamed Lophius Budegassa (2006); Froese and Pauly (2007); Empirical eq. 17 and 19 in Christensen et al. (2005) Kozul et al. (2001); Sinovcic et al. (2004); Vrgoc et al. (2006) Coll et al. (in press); Link et al. (2006) Froese and Pauly (2007), Stergiou and Karpouzi (2002) Source  0.0736 t·km-2  Notes  Species: Aidablennius sphynx, Coryphoblennius galerita, Lipophrys canevae, Parablennius zvonimiri, Parablennius sanguinolentus, Sarpa salpa, Parablennius gattorugine, Parablennius incognitus, Parablennius rouxi, Parablennius tentacularis, Salaria pavo, Chelon labrosus, Liza aurata, Liza Froese and Pauly ramado, Liza sapiens, Mugil cephalus, Tripterygion delaisi, (2007); Lipej et al. Tripterygion melanurus, Tripterygion tripteronotus, Lipophrys (2003) dalmatinus. Proportion between catches and biomasses; Z=M+Y/B (see Description of groups). Visual census data (using mean species weight and assuming habitat area in Eastern Adriatic from 0 to 100 m distance from coast = 437 km2)  P/B  Q/B  1.571 year  -1  13.193 year-1  Dulcic and Assumed GE=0.2 for L. aurata, L. ramada, L. saliens, M. Kraljevic (1997); cephalus, S. salpa; Inverted empirical eq. 19 in Christensen et Froese and Pauly al. (2005) (2007) Dulcic and Kraljevic (1997); Froese and Pauly Empirical eq. 17 in Christensen et al. (2005) (2007); Villamil et al. (2002)  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  GS  0.350  DC  Table 1  Coll et al. (in press); Link et al. (2006) Froese and Pauly (2007)  Zooplantivorous Value fishes  B  Source  15.922 t·km-2  Azzali et al. (2002); Cingolani et al. (2004a); Cingolani et al. (2004b); Froese and Pauly (2007); Lipej et al. (2003)  0.850 year-1  Q/B  8.776 year-1  GS  0.150  Coll et al. (in press); Link et al. (2006)  DC  Table 1  Froese and Pauly (2007), Stergiou and Karpouzi (2002)  Pelagic piscivorous fishes  Value  Source  3.311 t·km-2  Azzali et al. (2002); Froese and Pauly (2007); Jukic-Peladic et al. (2001); Ragonese et al. (n.d.)  -1  P/B  0.899 year  Q/B  6.379 year-1  GS  0.150  Notes Species: Atherina boyeri, Spicara maena, Spicara smaris, Sardina pilchardus, Sardinella aurita, Sprattus sprattus, Engraulis enchrasiculos, Trisopterus minutus, Chromis chromis, Oblada melanura, Spondyliosoma cantharus, Chelidonichthys lucernus. Acoustic surveys, Proportion  between catches and biomasses; Z=M+Y/B (see Description of groups). Visual census data (using mean species weight and assuming habitat area in Eastern Adriatic from 0 to 100 m distance from coast = 437 km2)  Bartulovic et al. Z=M+F or from literature; M from eq. B3 in Brey (1999) and (2004); Cingolani from literature. Inverted empirical eq. 19 in Christensen et al. et al. (2004a); (2005); assumed GE=0.12 for S. aurita and T. minutus, and Cingolani et al. GE=0.2 for C. lucernus (2004b); Dulcic and Kraljevic (1995); Dulcic et al. (2000); Dulcic et al. (2003); Froese and Pauly Empirical eq. 17 and 19 in Christensen et al. (2005) (2007); Pallaoro et al. (1998); Sinovcic (2001); Sinovcic et al. (2004)  P/B  B  21  Notes Species: Trachurus mediterraneus, Trachurus trachurus, Alosa fallax, Micromesistius poutassou, Scomber scombrus, Scomber japonicus. Acoustic surveys, MEDITS trawl surveys; Proportion between catches and biomasses; Z=M+Y/B (see description of groups).  Froese and Pauly Z=M+F or from literature; M from eq. B3 in Brey (1999) and (2007); Ragonese from literature. Inverted empirical eq. 19 in Christensen et al. (2005); assumed GE=0.2 for A. fallax and S. scombrus, and et al. (n.d.); GE=0.1 for M. poutassou Santic et al. (2002); Sinovcic et Empirical eq. 17 and 19 in Christensen et al. (2005) al. (2004) Coll et al. (in press); Link et al. (2006)  22  DC  Trophic Model of the Northern Adriatic Sea, Barausse, Duci, Mazzoldi, Artioli and Palmeri  Table 1  Zoobenthivorous Value fishes 1  EE  0.950  Froese and Pauly (2007) Source  Notes  Species: Callionymus lyra, Callionymus phaeton, Callionymus risso, Deltentosteus quadrimaculatus, Gobius bucchichi, Gobius cobitis, Gobius cruentatus, Gobius paganellus, Pomatoschistus marmoratus, Pomatoschistus minutus, Pomatoschistus norvegicus, Labrus merula, Symphodus Christensen et al. cinereus, Symphodus melops, Symphodus ocellatus, Symphodus roissali, Symphodus rostratus, Symphodus tinca, (2005) Mullus barbatus, Mullus surmuletus, Boops boops, Diplodus annularis, Diplodus puntazzo, Diplodus sargus, Diplodus vulgaris, Lithognathus mormyrus, Sparus aurata, Hippocampus guttulatus. Not enough reliable biomass estimates  P/B  0.776 year-1  Q/B  5.137 year-1  GS  0.300  DC  Table 1  Azevedo and Simas (2000); Curtis and Vincent Z=M+F; M from eq. B3 in Brey (1999). Inverted empirical eq. (2006); Dulcic and 19 in Christensen et al. (2005); Assumed GE=0.1 for G. Glamuzina (2006); cobitis, and GE=0.2 for G. paganellus, D. annularis and S. FAO Adriamed - aurata. Mullus Barbatus (2006); Froese and Pauly (2007); Kallianiotis et al. (2005); Kraljevic et al. (1996); Empirical eq. 17 and 19 in Christensen et al. (2005) Pallaoro and Jardas (2003); Stergiou and Karpouzi (2002) Coll et al. (in press); Link et al. (2006) Azevedo and Simas (2000); Froese and Pauly (2007); Kallianiotis et al. (2005); Stergiou and Karpouzi (2002)  Zoobenthivorous Value fishes 2  Source  EE  0.950  Species: Gaidropsarus mediterraneus, Gobius niger, Dicentrarchus labrax, Sciaena umbra, Umbrina cirrosa, Christensen et al. Scorpaena notata, Scorpaena porcus, Serranus hepatus, (2005) Serranus scriba, Pagellus erythrinus, Pagellus acarne, Syngnathus acus, Trigla lyra, Trigloporus lastoviza, Merlangius merlangus. Not enough reliable biomass estimates  0.648 year-1  FAO Adriamed Pagellus Z=M+F; M from eq. B3 in Brey (1999). Inverted empirical eq. Erythrinus (2006); 19 in Christensen et al. (2005); Assumed GE=0.3 for G. niger FAO Adriamed - and P. acarne, and GE=0.2 for D. labrax and S. umbra. Merlangius  P/B  Notes  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  Q/B  3.954 year-1  Merlangus (2006); Froese and Pauly (2007); JukicPeladic et al. (2001); Empirical eq. 17 and 19 in Christensen et al. (2005) Labropoulou et al. (1998); Spedicato et al. (2002); Zorica et al. (2006)  GS  0.300  Coll et al. (in press); Link et al. (2006)  23  Artuz (2005); Froese and Pauly (2007); Stergiou and Karpouzi (2002) Source Notes  DC  Table 1  European hake  Value  B  0.0475 t·km-2  P/B  1.157 year-1  Q/B  4.241 year-1  GS  0.150  Coll et al. (in press); Link et al. (2006)  DC  Table 1  Stergiou and Karpouzi (2002)  Rays  Value  Source  -2  B  0.0127 t·km  P/B  0.612 year-1  Q/B  3.058 year-1  GS  0.150  DC  Table 1  FAO Adriamed Merluccius Species: Merluccius merluccius. MEDITS trawl surveys Merluccius (2006) FAO Adriamed Merluccius Z=M+F; M from eq. B3 in Brey (1999). F=Y/B Merluccius (2006) FAO Adriamed Empirical eq. 19 in Christensen et al. (2005) Merluccius Merluccius (2006)  Notes  Species: Torpedo torpedo, Torpedo mormorata, Dasyatis Jukic-Peladic et al. centroura, Dasyatis pastinaca, Myliobatis aquila, Raja asterias, (2001) Raja clavata, Raja miraletus, Raja montagui. MEDITS trawl surveys. Assumed GE=0.2 (also, close to value from Coll et al., in press) Abdel Aziz (1992); Filiz and Bilge (2004); Froese Empirical eq. 17 in Christensen et al. (2005) and Pauly (2007); Ismen (2003); SIBM (2006) Coll et al. (in press); Link et al. (2006) Froese and Pauly (2007); Ismen (2003); Stergiou and Karpouzi (2002); Vannucci (2005)  24  Trophic Model of the Northern Adriatic Sea, Barausse, Duci, Mazzoldi, Artioli and Palmeri  Sharks  Value  Source  B  0.0262 t·km-2  Species: Mustelus mustelus, Scyliorhinus stellaris, Squalus Jukic-Peladic et al. blainvillei, Squalus acanthias, Mustelus asterias, Scyliorhinus (2001) canicula. MEDITS trawl surveys.  -1  Filiz and Mater (2002); Froese and Pauly (2007), Jukic-Peladic et al. (2001); Opitz (1996)  Notes  Z=M+F; M from eq. B3 in Brey (1999). Inverted empirical eq. 19 in Christensen et al. (2005). M divided by 2, because it is probably overestimated (Rodriguez-Cabello and Sanchez, 2005)  P/B  0.264 year  Q/B  2.877 year-1  GS  0.150  Coll et al. (in press); Link et al. (2006)  DC  Table 1  Cortés (1999); Froese and Pauly (2007)  Seabirds  Value  Source  B  0.0088 t·km-2  Matteo Griggio (Dept. Of Biology, Larus michaellis, Larus melanocephalus, Larus ridibundus, University of Larus canus, Puffinus yelkuoan, Calonectris diomedea, Padova) and Lorenzo Serra (Ist. Phalacrocorax carbo, Phalacrocorax aristotelis, Podiceps nigricollis, Podiceps cristatus. Naz. Fauna Selvatica) Pers. Comm.  P/B  4.610 year-1  Q/B  69.340 year-1  GS  0.125  DC  Table 1  Coll et al. (in press) Coll et al. (in press) Coll et al. (in press) Coll et al. (in press)  Empirical eq. 17 and 19 in Christensen et al. (2005)  Notes  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  25  UPDATED ECOSYSTEM MODEL FOR THE NORTHERN BENGUELA ECOSYSTEM, NAMIBIA1 Sheila J.J. Heymans  Scottish Association for Marine Science, Dunbeg, Oban PA37 1QA, Scotland; Email: Sheila.Heymans@sams.ac.uk  U. Rashid Sumaila  Fisheries Centre, 2202 Main Mall, Vancouver, BC, V6T1Z4, Canada; Email: r.sumaila@fisheries.ubc.ca  ABSTRACT An ecosystem model of the northern Benguela ecosystem from the coast to the 500 m depth contour (between 15°S and 29°S) was constructed for 1956. This model consists of 31 living compartments and detritus. It includes all the commercially important fish and marine mammal species in the ecosystem. The model was driven by effort time series for the 10 different fleets operating currently in the ecosystem, namely the purse seine, midwater trawlers and demersal fleets (all fishing from 1956 to 2003); the longline tuna fleet fishing from 1961 to 2003; the deep water crab fleet (1973–2003); the lobster fleet (1956–2003); commercial and recreational line fishery (1956–2003); seal hunt (1956–2003) and the seaweed collection fishery (1980–2003). The model was fitted by changing the vulnerabilities of predators to their prey and by estimating an environmental anomaly that was significantly correlated with sea surface temperature (negative correlation) and wind stress (positive correlation). The temporal model reproduced the general anchovy decline but was not able to reproduce the large increase in sardine in 1960 that was estimated by Virtual Population Analysis (VPA). The good fit of the hake and monkfish biomass in the model to the data was related to the high catches in the system, and their catches were reproduced very well, as were the catches of other demersal fish species and crabs. For predators such as seals, sharks and snoek the model reproduced the catch series well even when no good biomass estimates were available. The spatial model reproduced the known distribution of benthic species such as crabs and lobster as well as more mobile species such as seals, snoek, hake and others.  INTRODUCTION Ecosystem models of the northern Benguela include those constructed by Heymans (1997), Shannon and Jarre-Teichmann (1999), Heymans and Baird (2000) and Roux and Shannon (2004) and span the time period from 1970 to 1999. Heymans (2004) examined the effects of internal (bottom up) and external (top down) control by means of ecosystem modelling from the 1970s onwards. Similarly, Cury and Shannon (2004) documented top-down control (fishing) and bottom-up control that initiated and sustained regime shifts or species replacements via environmental forcing. The aim of this model developement is to look at the northern Benguela over a longer time span and to model the system from 1956 onwards. A secondary aim is to look at the spatial patterns within the ecosystem. By 1956 the main fisheries were either underway or just beginning. This report therefore describes an Ecopath model for 1956 where possible or for the 1950s in general, all the time series data available since then to reproduce these trends in Ecosim, and the spatial information available to construct an Ecospace model of the ecosystem.  1 Cite as: Heymans, S.J.J. and Sumaila, U.R. 2007. Updated ecosystem model for the northern Benguela ecosystem, Namibia, p. 25– 70. In: Le Quesne, W.J.F., Arreguín-Sánchez, F. and Heymans, S.J.J. (eds.) INCOFISH ecosystem models: transiting from Ecopath to Ecospace. Fisheries Centre Research Reports 15(6). Fisheries Centre, University of British Columbia [ISSN 1198-6727].  26  Northern Benguela Ecosystem, Heymans and Sumaila  MATERIALS AND METHODS Spatially the ecosystem model of the northern Benguela ecosystem extends between 15°S and 29°S from the coast to the 500 m depth contour, a total area of approximately 179,000 km². The model consists of 31 living compartments and detritus. The living compartments include 2 marine mammal groups, seabirds, 18 fish groups, 8 invertebrate groups and 2 primary producers. It includes all the commercially important fish, of which 6 species were divided between adults and juveniles, namely anchovy (Engraulis japonicus), sardine (Sardinops ocellatus), gobies (Suffoglobius bibarbatus), horse mackerel (Trachurus trachurus capensis), hake (Merluccius spp.) and jellyfish (mainly Chrysoara hyoscella and Aequorea aequorea). The main fisheries in the system are also defined, with economic and social information added for policy exploration.  Fishery Best (2006) gives data on the French whaling fleet of the late 1700s, showing that French whalers fished in Walvis Bay and Tiger Bay (northern Namibia) from 1787 to 1793 at least. The English started whaling off the coast of Namibia in the nineteenth century and exploited guano from the offshore islands; by the latter part of the nineteenth century guano mining, sealing and fishing establishments at Sandwich Harbour were major commercial enterprises on the Namibian coast (Kinahan, 1991). Whaling off Namibia usually occurred from May to August (Best, 2006). The first land-based commercial fishery in Namibia was established at Sandwich Harbour in 1851 where fish were caught with hook and line from rowboats, salted and sun dried, then shipped to South Africa, from where they were exported to Mauritius (Kinahan, 1991). The fishery caught silver kob (Argyrosomus spp.), steenbras (Lithognathus aureti), sea barbel (Galeichthys feliceps), hake and snoek (Thyrsites atun), and also exported fish oil that consisted of shark liver oil (Elasmobranchii), whale (various species) and seal blubber (Cape fur seal, Arctcephalus pusillus pusillus) (Kinahan, 1991). The natural closing of Sandwich Harbour in 1891 ended the fishery, and no commercial fishing took place for the next 73 years (Holtzhausen et al., 2001). According to Sparks (1984), most fishing before World War II was small scale and seasonal, and the industry only expanded after the war. Canning activities and fish oil production began in the 1950s and were based in Walvis Bay (Sparks, 1984). The South African fishers were being restricted in their own country and began moving into Namibian waters, and by the 1960s the catches had expanded rapidly (Sparks, 1984). By the mid-1960s as many as 100 foreign vessels, under 15 or more national flags including the USSR, Poland and Spain, were fishing off Namibia (Fuller and Prommer, 2000). The Ukrainian fleet started fishing in the northern Benguela in 1963 and initially targeted sardine and Cape horse mackerel, but in 1968 hake became one of the principal target species (Romanov, 2001). The fleet switched between horse mackerel and hake depending on fishing success, but the fishery for sardine was stopped early (Romanov, 2001). In the mid-1980s hake biomass declined and regulatory measures for hake were introduced; the Ukrainian fleet shifted to chub mackerel (Scomber japonicus) (Romanov, 2001). Ten fisheries were defined in the model to represent the dynamics of the fishery operations in the region from 1956 to 2003, namely: 1. Purse seine fishery The pelagic purse seine fishery targeted sardine since before 1947 (Hampton, 2003) and anchovy since 1964. Lees (1969) found that in 1952 there were over 100 purse seine boats and 6 factories operational in Walvis Bay. Similarly, in 1953 there were 6 factories in Walvis Bay and Lüderitz and 100 small purseseiners were catching 262,000 tonnes of sardine (Hampton, 2003), which were exported in 509,234 cartons of canned sardines (Lees, 1969). The factories had a quota of 90,000 tonnes each (Lees, 1969). By 1955 there were 217 boats, and in 1956 around 1.8 million cartons of canned sardines were exported (Lees, 1969). The pelagic fishery has always been situated around Walvis Bay and Lüderitz, and currently they fish mainly for juvenile horse mackerel, which is situated in the north (Hampton, 2003). Therefore the main ports of fishing for this fleet in the spatial model were Lüderitz and Walvis Bay.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  27  In 1960, Hart and Currie (1960) reported that the pelagic fishes were “already being increasingly exploited and among them the South African pilchard takes first place as the basis of extensive fisheries at various points along the coast. Following closely in importance are fisheries for the ‘maasbanker’ (Trachurus trachurus), stockfish (Merluccius capensis) and snoek (Thyrsites atun), and many sharks are taken principally for their liver oil”. In the early 1960s, the pilchard industry changed to trawling and small-scale line fishing, and two factories started freezing and salting fish for sale and export (Holtzhausen et al., 2001). The canneries increased from 5 to 6 in 1961 and exported 4.7 million cartons of sardines that year (Lees, 1969). In 1963 the factories increased to 8 and in 1964 the fishery experimented with catching anchovy; the 7 boats that fished specifically for anchovy caught only 718 tonnes, while 4.5 million cartons of sardine were exported (Lees, 1969). By 1967 the quotas of the factories increased from 90,000 tonnes to 99,600 tonnes each, and two factory ships were also fishing off Namibia. In 1968 the factory quotas increased to 120,000 tonnes each, the factory ships had a combined catch of 630,000 tonnes and the Russian vessels took 300,000 tonnes of sardine (Lees, 1969). Since the decline of these species in the early 1980s, the pelagic fishery has concentrated on juvenile horse mackerel, and operating with a mesh size between 11 and 23 mm they usually catch horse mackerel less than one year old (Vaske, 2001). Since 1990, the purse seine fishery is entirely Namibian-owned and consists of steel- and wooden-hulled vessels that operate out of Walvis Bay from mid-February to the end of August (Hampton, 2003). In 2000 the fleet consisted of 30 registered vessels between 21 and 49 m (99–614 tonnes), but by 2001 only 14 of these vessels were operational due to scarce fish (Hampton, 2003). The fleet has declined from 35 to 45 vessels in the 1980s to fewer than 15 vessels by 2000 in response to declining total allowable catches (TACs) (Boyer and Oelofsen, 2004). In 1999 there were still 40 vessels in operation, and that number had decreased to 13 by 2002 (Boyer and Oelofsen, 2004). These vessels are active for just a few months of the year and catch less than 10 tonnes per GRT (Gross Registered Tonnes) each year, while in the mid 1970s they caught 90 tonnes per GRT (Boyer and Oelofsen, 2004). Also, most of the Namibian canning and fishmeal plants are idle for much of the year (Boyer and Oelofsen, 2004). If the purse seine fleet catches more than 5% juvenile sardine, the area is closed to fishing for several weeks, although discarding of young sardine does take place (Boyer and Oelofsen, 2004). With the rights to exploit small pelagic fish, limited quotas for sardine and juvenile horse mackerel are issued, but there are no limitations on how much anchovy, round herring, chub mackerel and pelagic goby (Suffoglobius bibarbatus) is caught (Boyer and Oelofsen, 2004). Effort time series for the purse seine fishery was estimated from the fleet hold capacity obtained for the fleet from 1956 to 1975 from Fuller and Prommer (2000), and the effort obtained from 1975 to 1987 from le Clus et al. (1988) (Figure 1).For 1988 to 2003 the effort was estimated to be related to the catch by using the average effort-to-catch ratio for 1980 to 1987, and the estimated effort showed a similar trend to the TACs set for the fleet from 1990 to 2001 (Nicols, 2004). Pu rse seine, demersal  Klingelhoeffer (2006) suggests that the midwater trawl fishery started in the 1960s and catches mostly horse mackerel (Bauleth-D’Almeida et al., 2001), which is mostly situated in the northern part of the ecosystem. However, no specification was made on the ports for this fishery and fishing is allowed anywhere in the spatial model.  300  30  200  20 100 10  Midwater, longline  40  2. Midwater trawl fishery  0 0 1956 1961 1966 1971 1976 1981 1986 1991 1996 2001 Purse seine  Demersal  Midwater  Longline  The midwater trawl fleet uses nets of 60 mm mesh to target adults Figure 1. Effort time series (relative to 1956) used to drive the horse mackerel (17–48 cm), which northern Benguela model for the purse seine, demersal, midwater and occur offshore of the 200 m isobath, longline fleets. while the purse seine fleet uses nets of 12 mm mesh to catch juvenile horse mackerel of 6–20 cm inshore of the 200 m isobath (BaulethD’Almeida et al., 2001). By 2001 the midwater fleet consisted of 25 large vessels from 75 m to 120 m with a  28  Northern Benguela Ecosystem, Heymans and Sumaila  maximum GRT of 7,765 tonnes; they process the catches at sea (Hampton, 2003). They are operational throughout the year. Most of the catch is frozen and transshipped to reefer vessels for export as a relatively low value product to West Africa, while 10% is reduced to fish meal and 5% is dried and salted ashore for export to African countries (Hampton, 2003). Bycatch of other species makes up about 2% of the catch of horse mackerel (Hampton, 2003). If catches of hake or young horse mackerel exceed 5% of the catch the fleet must leave the area (Boyer and Oelofsen, 2004). To get an effort time series for the midwater trawl fishery, some calculations and assumptions had to be made. From the FAO catch statistics (FishStat), the nations fishing for Cape horse mackerel in 1986–1989 included South Africa, USSR, Cuba, Bulgaria, Romania, Japan, Israel, Angola, DRC, Poland, Spain, Portugal, Iraq and Namibia, while in 1990 the nations included South Africa, Russia (former USSR), Cuba, Bulgaria, Romania, Japan, Israel, Germany, Latvia, Lithuania, Estonia, Georgia, Ukraine and Namibia. In addition, the number of boats fishing from 1990 to 2004 was given by Klingelhoeffer (2006) starting at 100+ in 1990 down to 24 in 2004, while the effort by the most important fleets (USSR, Poland, Romania and Bulgaria) for 1973–1986 was given by Butterworth et al. (1990). Thus if one assumes that because there were the same number of nations fishing from 1986 to 1990, the number of boats in 1986 would be similar to the 100+ boats in 1989, and if their standardized effort did not change much over that time, then the number of boats can be calculated back to 1973 from the effort. In addition, if one uses total catch instead of effort, the number of boats is not significantly different; therefore, for 1955–1972 the same calculation was used but using catch instead of effort to back calculate the effort (Figure 1). This gives a very similar number of boats as the effort calculation for 1972–1986. However, to get an effort of 1 in 1956, the number of boats in 1986–1990 has to be increased to 130. 3. Demersal fishery Klingelhoeffer (2006) suggests that a demersal trawl fishery was initiated in the 1950s and that the demersal trawl fishery targeting hake was dominated by Spain (Klingelhoeffer, 2006). By 1965 one hake processing plant was operational in Walvis Bay and by 1967, this had increased to 3 processing plants (Lees, 1969). On the shelf, directed bottom trawling is carried out for hake, monkfish (Lophius spp.) and sole (Austroglossus microlepis) by a fleet of Namibian registered freezer and wet-fish trawlers based in Walvis Bay or Lüderitz, varying from 20 to 74 m in length and from 84 to 1,780 GRT (Stuttaford 1999 in Hampton, 2003). There were 111 such vessels in 2000. The larger vessels are capable of trawling to depths off the edge of the shelf or greater than 500 m (Hampton, 2003). The hake vessels operate throughout the year between the legal minimum depth of 200 m and the shelf edge and fish both by day and (less frequently) by night (Hampton, 2003). As the fleet includes both freezer and factory trawlers (Van der Westhuizen, 2001), it was assumed to fish over the whole area of the spatial model using the ports of Walvis Bay and Lüderitz for operational purposes. Hake are principally caught by bottom trawling freezer/factory and wet-fish trawlers, but they are also caught by longlines and as bycatch in the monkfish and sole fishery and by the horse mackerel fishery (Van der Westhuizen, 2001). The hake fishery is the major contributor to employment in the fishing sector, and by 2000 there were about 4,500 shore-based and 2,500 sea-based employees in the hake fishery (Van der Westhuizen, 2001). The sector’s contribution to the GDP was 1.7% in 1990 and more than 10% in 1999 (Van der Westhuizen, 2001). The most abundant bycatch species in the hake-directed trawls are monkfish, kingklip, horse mackerel and snoek (Anon., 2001 in Hampton, 2003). Bottom trawlers used for hake fishing were used to estimate the percentage composition by weight of the main demersal taxa by survey between 1992 and 1996 (Hamukuaya et al., 2001). This study found that on average 95% of the catch made by the demersal trawl consisted of teleosts, 3% of chondrichthyans and 2% of invertebrates, which included cephalopods, crabs, stomatopods, shrimps and lobsters (Hamukuaya et al., 2001). These ratios were therefore used to estimate the catch of sharks and invertebrates from the trawl fleet. The breakdown of the teleosts and chondrichthyans show that cape hake dominate both the shelf and slope assemblages, but that shallow water hake (Merluccius capensis), horse mackerel and the goby dominate the shelf and upper slope area where the oxygen content is often low (Hamukuaya et al., 2001).  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  29  The monkfish and sole vessels are generally smaller than the hake trawlers and do not operate as far from the coast, although they also fish throughout the year (Hampton, 2003). These high value fish are mainly exported to Europe in various frozen forms and the most abundant bycatch species in the trawls directed at monkfish and sole is hake (Hampton, 2003). In 1994 an experimental license was given to one fishing company to fish in water exceeding 700 m for monkfish, deep water catfish (Lepidion capensis), warty dory (Allocyttus verrucosus), spiky dory (Neocyttus rhomboidalis), alfonsino (Beryx splendens), jewel squid (Histioteuthis spp.) and trachichthyids (Hoplostethus spp.) (Boyer et al., 2001b). Viable fishing populations of orange roughy (Hoplostethus atlanticus) were found at 4 fishing grounds separated by 200 km (Boyer et al., 2001b). The vessels that fish for orange roughy and alfonsino vary in length between 28 and 55 m, and they trawl down to depths of 600–900 m with heavy-duty trawl gear (Hampton, 2003). This fishery is too deep to be in this model, as the fish are not included in this ecosystem. No effort time series was available for the demersal fleet, but the number of boats catching hake from 1991 to 2000 was obtained from Van der Westhuizen (2001) and showed a linear relationship with the biomass in that time period. Thus it can be assumed that the effort was linearly related to biomass for the whole time period (Figure 1). 4. Longline fishery The most important longline fisheries for large pelagic species are those for tuna, particularly longfin or albacore tuna (Thunnus alalunga) and bigeye tuna (T. obesus), swordfish (Xiphias gladius) and large pelagic sharks. These species are caught by a fleet of Namibian and foreign bait-and-pole vessels and by foreign longliners. Most of the bait boat catches are made in the extreme south of Namibia, in contrast to the longline catches, which are more widely spread latitudinally and are generally further offshore, often outside of the Namibian exclusive economic zone (EEZ) (Hampton, 2003). Japanese and Taiwanese longliners have been fishing for tuna off Southern Africa since the early 1950s (Ryan et al., 2002). In general the longline fleet is widespread and further offshore than the bait boats, which catch mainly in the south (Hampton, 2003). Thus, the fleet was operational throughout the whole spatial model area. Hake are caught by 24 Namibian-owned deep longliners that operate from Walvis Bay and Lüderitz throughout the year (Hampton, 2003). The hake longline fishery started in the mid-1980s (Hampton, 2003). These catches are processed and exported as fresh fish to the lucrative European market (Hampton, 2003). Between 5,000 and 10,000 tonnes of hake are caught annually by the longliners (Hampton, 2003). The longliners have little effect on the sea bottom, but the effect on seabirds such as gannets and albatrosses is severe (Hampton, 2003). Tuna is caught by approximately 30 local and foreign-owned pole or longline vessels (Hampton, 2003). The rights for tuna and swordfish fishery have recently been extended to include all large pelagics, and as there are no TACs for sharks they are being targeted and several thousand tonnes are now caught annually (Boyer and Oelofsen, 2004). As no estimate of effort was available for the tuna longline vessels and we have effort only for 1998–2003 for the hake longline vessels, we assumed that effort was related to catch for the longline vessels and specifically to the tuna catch as the hake catch was comparatively small (Figure 1). This is supported as, for the 5 years that the number of hake vessels are available (Nicols, 2004) it shows the same trend as the effort estimated from the tuna-catch derived effort. 5. Crab fishery The red crab (Chaceon maritae) fishery started in 1973 and by 1974 there were 17 Japanese vessels and one mother vessel of 1,500 gross tonnes targeting this species (Beyers and Wilke, 1980). Effort declined due to marketing problems and by 1979 only 5 vessels were operating (Beyers and Wilke, 1980). By 2001 the number of vessels targeting this species had declined to 2 (Hampton, 2003). Beehive-type traps are used on a longline, with each vessel carrying 1,200 traps (Beyers and Wilke, 1980). The entire red crab catch is processed at sea and exported to Japan (Hampton, 2003). Crabs are caught in the north (Hampton, 2003) and are usually found in deep water with their main area of capture between 18°S and 21°S (Le Roux, 2001). The number of crab vessels per year from 1973 to 1986 was obtained from MelvilleSmith (1988) and from 1998 to 2003 from Nicols (2004). A straight linear projection was made between the 5 boats in 1986 and the 3 boats in 1998 to give an effort series for this fleet (Figure 2).  Northern Benguela Ecosystem, Heymans and Sumaila  Lobster effort time series in trap-days*103 were obtained from Grobler and Noli-Peard (1997) for 1958 to 1996; we assumed that the effort in 1956–1957 was similar to that of 1958 (Figure 3). Pulfrich et al. (2003) give estimates of catch per unit effort (CPUE) from 1995 to 1999, which were used in conjunction with the catch to estimate an effort that was linearly scaled to those in Grobler and Noli-Peard (1997). For 2000–2003 the effort was assumed to be correlated with the catch and scaled accordingly.  6  18  5  15  4  12  3  9  2  6  1  3  0 0 1956 1961 1966 1971 1976 1981 1986 1991 1996 2001 Crab  Seals  Seaweed  Figure 2. Effort (relative to the first year) with catch in the crab, seal and seaweed fisheries.  10 Relative fishing effort  The lobster (Jasus lalandii) fishery started in the 1930s. Rock lobster is caught in summer on shallow reefs off southern Namibia and taken in hoop nets from dinghies and in lobster traps set by 20 larger vessels (Hampton, 2003). Most of the lobster is cooked and exported to Japan (Hampton, 2003). The hub of the lobster fishery is Lüderitz, which was specified as the port of landing for this fishery in the spatial model.  Crab, seals  6. Lobster fishery  Seaweed  30  8 6 4 2 0 1956 1961 1966 1971 1976 1981 1986 1991 1996 2001 Lobster  Commercial linefish  Recreational line  Figure 3. Effort (relative to 1956) in the lobster, commercial and recreational line fishery.  7. Commercial line fishery The commercial line fishery uses rod–and–reels or hand lines with baited hooks for recreational fishers from the shore or from ski boats, or commercially from ski boats and line boats (Holtzhausen and Kirchner, 2004). Line boats are approximately 20 m long and carry up to 16 fishers using hand lines with two hooks each; they catch mostly kob, steenbras and snoek (Holtzhausen and Kirchner, 2004). By 2002 there were 20 firms registered as permit holders in the line fishing industry of which 7 were ski boat operators and 13 larger vessels (Stage and Kirchner, 2005). The commercial line fishery operates about 10 vessels in inshore waters up and down the coast from Walvis Bay (Zeybrandt and Barnes, 2001). The first catches from the line boats were recorded in 1964 for kob and in 1973 for steenbras (Holtzhauzen (1999) and Krichner (1998) in Holtzhausen et al., 2001). Ski boats are approximately 5–6 m long and carry between 4 and 6 fishers who work with one rod and reel each. They do not usually catch steenbras, but they catch all other species, while shore anglers do not usually catch snoek, but also catch most everything else (Holtzhausen and Kirchner, 2004). As we do not have any estimate of the commercial line fishery effort, it was assumed to be related to the catch of all kob, steenbras and snoek in the model (Figure 3). The main areas that are open to the line fishery include the West Coast Recreational Area (WCRA), which stretches from the Ugab River mouth at 21°S to Sandwich Harbour at 23.5°S, with some fishing being possible in the Skeleton Coast Park north of the WCRA; however, no fishing is possible in the Namib Naukluft Park and Diamond Areas to the south (Holtzhausen and Kirchner, 2004).  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  31  8. Recreational line fishery The most important recreational angling fish off Namibia are kob, West Coast steenbras, galjoen (Dichistius capensis), blacktail (Diplodus sargus) and various shark species such as the spotted gully shark (Triakis megalopterus), coppershark (Carcharhinus brachyurus), cowshark (Notorynchus cepedianus) and smooth houndshark (Mustelus mustelus) (Holtzhausen and Kirchner, 2004). Shore angling occurs along approximately 260 km of the Namibian coastline, between Sandwich Harbour to the mouth of the Ugab River (Zeybrandt and Barnes, 2001). Over 90% of the angling taking place in the WCRA in the vicinity of Walvis Bay, Swakopmund and Henties Bay (Stage and Kirchner, 2005), although there are also small sites north at Torra Bay and Terrace Bay in the Skeleton Coast Park and south near Lüderitz (Zeybrandt and Barnes, 2001). A small proportion of angling (<5% of effort and <3% of catches) is done for subsistence (Kirchner et al., 2000). In the spatial model the main recreational fishing took place in the WCRA, at Torra Bay and Terrace Bay in the north and at Lüderitz in the south. According to Holtzhausen et al. (2001), recreational fishery along the Namibian shoreline was anecdotally quite good prior to 1990, but no catch or effort data was available for the fishery except for the estimates obtained from Holtzhausen and Kirchner (2004) that indicated that the angling catch was about one third of the commercial catch for kob and steenbras. The effort of the recreational fishery (Figure 3) was assumed to be similar to the catch for all linefish species estimated by the Sea Around Us (www.seaaroundus.org) and other sources (see linefish below). The recreational shore angler pays most per fish and is therefore most valuable to the Namibian economy, while the commercial line fishery is worth only 1/7th the value of recreational fishery (Holtzhausen et al., 2001). Barnes et al. (2004) studied the economic value of the recreational fishery and found that approximately 8,300 anglers spend a total of 173,000 days angling, and each angler spends about N$3,400 for a total expenditure between N$23 million and N$31 million or 2–4% of the total fishing sector. 9. Seal fishery Seal harvesting has taken place in southern Africa since the seventeenth century, and legal controls were implemented in 1893 (Butterworth et al., 1995). The first sealing company in Namibia started in 1895 at Cape Cross and by the end of that year the company had shipped 70 workers from Britain, Germany and South Africa (Berry, 2002). The company, Damaraland Guano Company, started off collecting guano and expanded into harvesting seals soon after (Berry, 2002). Data was available on the pup harvest since 1900 (David, 1989) and bulls since 1901 (Butterworth et al., 1995). Winter harvesting is focused on the yearlings and during summer it is limited to the bulls (Hart and Currie, 1960). Cows were only harvested in the late 1980s and early 1990s (Butterworth et al., 1995). Legislation controlling the utilization of seals in Namibia was introduced in 1922 (Berry, 2002). Seal pelts and genitalia are exported, while oil and meat are also produced (Wickens et al., 1991). The seal fishery occurs mainly at Cape Cross (21.5°S) (Wickens et al., 1991) and was entered into the spatial model this way. No effort was available for the seal fishery, and it was assumed to be related to the catch of seals (Figure 2). 10. Seaweed harvest Formal seaweed harvesting started in 1980 (Hampton, 2003), although prior to that informal harvesting was taking place. Beach-cast Gracilaria are collected and supplemented with Gracilaria growing on longlines (Hampton, 2003). A company was established in Lüderitz in 1992 that can process up to 6,000 tonnes of seaweed annually to the agar stage, and good prices are obtained on the world market because the gel strength is very high (Hampton, 2003). The company, Taurus Atlantic Seaweeds, cultures Gracilaria in a 40 ha plot in the Lüderitz lagoon (Anon., 2005). Thus the seaweed fishery only occurs in the Lüderitz bay area of the spatial model (Rotmann, 1987; Critchley et al., 1991; Hampton, 2003). The annual production of Gracilaria is around 120 tonnes of dry weight per year, and the operation employs 50 people (Anon., 2005). The effort of this fishery was assumed to be related to the catch for the fishery (Figure 2).  Management In June 1949 the South West African Administration, the government of what was then South West Africa, passed the Sealing and Fishing Ordinance, which gave the Administration considerable power over all fishing activities (Lees, 1969). It specified the maximum quantities of any specific fish species that could  32  Northern Benguela Ecosystem, Heymans and Sumaila  be treated in the factories, and it limited the number of fishmeal and oil reduction plants and floating processing factories (Lees, 1969). It provided for taxes on fish and fish products, closed seasons and sanctuaries and collected statistics from fishermen and factory owners (Lees, 1969). A laboratory was built in Lüderitz but closed down quickly, and by the 1960s research was carried out by the Marine Research Laboratory in Walvis Bay (Lees, 1969). In 1950 the first scientist was appointed and in 1952 the first research vessel, the Namib II, was launched (Lees, 1969). For the first few years anyone with a licensed boat could catch fish, and there were no restrictions on the tonnage landed, so that by 1952 there were 11 industrial sites in Walvis Bay and the fourth major factory was being built (Lees, 1969). There were also canneries in Lüderitz mainly for the canning of lobster (Lees, 1969). By the late 1960s there were 8 factories processing pelagic fish in Walvis Bay and 3 processing hake (Lees, 1969). The factories had a limit of 90,000 tonnes of pelagic fish until 1967 when the limit was increased to 99,600 tonnes and in 1968 to 120,000 tonnes (Lees, 1969). During that time 2 processing ships were also fishing off Namibia, and their TAC was limited to 570,000 short tons in 1969 (~520,000 tonnes) and 500,000 short tons (~450,000 tonnes) in 1970 (Lees, 1969). The Division of Sea Fisheries in Cape Town assumed control of the South West African research organization in 1969 (Lees, 1969). From the 1970s to 1990 horse mackerel (and hake) was assessed by the International Commission for Southeast Atlantic Fisheries (ICSEAF) through the Standing Committee on Stock Assessment (Anon., 2001). The Ministry of Fisheries and Marine Resources of Namibia took over the assessment in 1990 (Anon., 2001). Scientists from the Ministry of Fisheries and Marine Resources based their recommendations on acoustic surveys and age- and length-based VPA estimates from commercial data (Maurihungirire, 2004). Under ICSEAF a TAC for horse mackerel was introduced in 1980, intensifying fishing for this species, and regulatory measures were only imposed on the foreign fleet while unrestricted catches were allowed by the local purse seine fleet (Anon., 2001). The only limiting factor on the local fleet was the closed fishing season in August (Anon., 2001). The TAC was estimated based on an age-based VPA model (BaulethD’Almeida, 2001). Since Namibian independence in 1990, fishing is limited through output controls that consist of individual non-transferable quotas, and catches must be landed at one of two fishing ports while transshipments at sea are not permitted (Boyer and Boyer, 2004). Catches are landed at one of two fishing ports under the control of fisheries inspectors, and patrol vessels, aerial patrols and on-board observers ensure that the legal requirements are met (Boyer and Boyer, 2004). Ninety percent of the landings are from TACcontrolled stocks and therefore the Namibian authorities can control the fishing pressure of these stocks (Boyer and Boyer, 2004). Scientific personnel from the Ministry of Fisheries and Marine Resources base their recommendations on acoustic and midwater trawl surveys, with TACs being recommended for sardine at 18% of the survey biomass at the end of the previous fishing season (Maurihungirire, 2004). In Namibia, no edible or marketable fish taken as bycatch may be discarded, and this is monitored by shipboard observers, with levies being paid on the bycatch that would discourage targeting of such species (Boyer and Boyer, 2004). Subsequent to 1990, when Namibia declared their EEZ, a fishing mortality of F=0.3 year-1 was instituted for horse mackerel, which corresponded to an exploitation rate of 26% (C/B=0.26) (Anon., 2001). A quota was allocated to the pelagic fishery to prevent the removal of large quantities of juvenile horse mackerel, while bycatch, size and depth restrictions were instituted (Anon., 2001). Limitations were introduced on 1) bycatch to prevent high catches of juvenile hake and pilchard, 2) size and 3) depth, with restrictions of 200 m to curb high catches of juvenile hake, pilchard and juvenile horse mackerel (Bauleth-D’Almeida, 2001). It was proposed that the midwater fleet vacate an area whenever the proportion of horse mackerel <17 cm total length (TL) in a haul exceeded 5% by weight and high catches (5% per set) of horse mackerel <12 cm were discouraged (Bauleth-D’Almeida, 2001). For hake, the TAC is based on trends in abundance during fishery-independent bottom trawl surveys while making adjustments to allow for fish off the bottom at night, as determined acoustically, with the TAC set at 20% of the estimated biomass of mature hake (Maurihungirire, 2004). In 1998 an Interim Management Procedure was implemented and the TAC was adjusted according to the mean change in the survey and catch rate indices for the previous 5 years (Butterworth and Geromont, 2001; Maurihungirire, 2004).  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  33  Red crab management is based on length-based cohort analysis and predictive models, with TACs recommended on the basis of a projection of future biomass of the stock as a function of the catch (Maurihungirire, 2004). Namibia has developed a management plan for orange roughy, operational management plans for hake and seals and national plans of action (NPOA) are being drafted for seabirds and sharks (Boyer and Boyer, 2004). Illegal fishing within Namibia’s EEZ is believed to have been virtually eliminated as only one incident has been recorded since 1995 (MFMR 2002a in Boyer and Boyer, 2004).  Economics Fish exports from Namibia have increased from about N$830 million in 1992 to N$2,6 billion in 2000 (Boyer and Boyer, 2004). The value of the major industrial fisheries in Namibia in 2000 was given by Sumaila et al. (2002). In general, prices for species were obtained from Sumaila et al. (2007). Prices for sharks caught by longlines and the commercial line fishery were assumed to be similar to those obtained by the demersal trawl fishery, while sharks caught by the recreational fishery had higher prices. Recreational linefish prices were estimated based on the catch of linefish caught by the recreational fishery (approximately 1000 tonnes) times the value of the catch, approximately N$7 million, giving a price of N$12,363.59 per tonne. This price was also used for sharks caught by the recreational fishery. Tuna caught by the purse seine fleet was assumed to have the same price as those caught by the bait fishery. The export price per tonne for seaweed was US$1,250 in 1994. Prices for seals were obtained for 2000 from Hugo (2006) and all prices are given in Table 1. Table 1. Landed value in Namibian dollars (N$) for the different species caught in the northern Benguela by fleet. Group Name Seals Sharks Tuna Snoek Other linefish Anchovy Sardine Gobies Other s. pelagics Mesopelagics Juv. h mackerel Adult h mackerel Juv. hake Adult hake Monkfish Other demersals Crabs Lobster Benthic producers  Purse seine  Midwater trawler  Demersal  Long lines  Crab traps  Lobster  Com. line fishery  Recr. line fishery  6,597 7,479 6,597 6,597  12,364  Seal -ing  Other  621 6,597 7,479  6,597 24,150  3,158  12,364  371 834 371 371 371 397 2,065 3,158 5,685 7,220 6,597 11,357 71,700 1,028  In terms of product value the fishery for kob and other angling species in Namibia is insignificant compared with the major commercial fisheries (Hampton, 2003). However, the recreational fishery is an important tourist attraction and generates considerable expenditure far in excess of the product value. Kirchner et al. (2000) estimated that between October 1997 and September 1998, some 8,800 anglers spent 173,000 days angling and had direct expenditures totaling almost N$30 million. Value added to gross national income within the shore-angling fishery during that period was estimated at N$14 million (Hampton, 2003). Foreign visitors contributed 55% of the expenditure, and ski boat fishermen contributed about N$2 million annually to Namibia’s GDP, while the line boat fishery contributed about N$3.4 million (Kirchner, 1998 in Holtzhausen et al., 2001). The total contribution by the line fishery to the GDP was N$35 million. The total export value of trawled hake in 2000 was N$1.58 billion, more than the N$837 million in 1998, and 62% of this value was obtained from wet-fish hake and 38% from hake frozen at sea (Hampton, 2003). Monkfish exports in 2000 amounted to N$158.8 million, which is 10% more than in 1999. In 2000 horse mackerel exports from the midwater fishery were valued at N$596 million, and approximately 3% of  34  Northern Benguela Ecosystem, Heymans and Sumaila  the catch is consumed in Namibia (Hampton, 2003). The purse seine fishery exported N$117 million worth of sardine in 2000 compared to the N$320 million in 1998 (Hampton, 2003). Similarly, the export of horse mackerel from the pelagic fishery also declined from N$60 million in 1998 to N$15.8 million in 2000 (Hampton, 2003). The rock lobster trap fishery contributed N$34 million in 2000 and N$26 million in 1999 (Hampton, 2003). The rock lobster fishery is important as it is a relatively low capital, labour intensive industry that provides employment in a part of the country that has very high unemployment (Hampton, 2003). The crab fishery amounted to N$35 million in 2000 and N$25.5 million in 1999 (Hampton, 2003). The value of all tuna exported from Namibia in 2000 was N$29 million compared to the N$16 million in 1999 (Hampton, 2003). Seaweed harvesting started in 1980 and in 2000 829 tonnes were harvested with an export value of N$3,850,000 (Hampton, 2003). Profit and loss was estimated from data obtained from the annual fisheries income and expenditure report from the Ministry of Fisheries and Marine Resources for 1997, where the net profit divided by the total income from the fishery gave the percentage profit for each year (see Table 2 below). The profit for the seaweed industry was estimated at 13% by Rotmann (1987). For recreational line fisheries the profitability was estimated from the income of N$47.9 million and the profit of N$23.9 million obtained from Stage and Kirchner (2005). For seals and tuna we assumed a high profit margin of 40% (Table 2). The number of jobs Table 2. Income, expenditure, profit and percentage profit for the different fleets in that each fleet Namibia (Namibian $). Income Expenditure Net Profit/Loss % profit supports was Fishery 309,327,946 319,638,333 -10,310,387 -3 estimated from the Purse seine 242,779,080 219,722,050 23,057,030 9 Ministry of Fisheries Midwater 756,066,479 795,463,632 -39,397,153 -5 and Marine Resources Demersal 40 for the purse seine, Longline tuna* Crab 254,809,630 207,414,600 47,395,030 19 midwater, demersal, Lobster 26,896,000 22,514,520 4,381,480 16 crab, lobster and 1,593,740 1,237,870 355,870 22 commercial line Commercial line fishery Recreational line 50 fisheries, while the 40 data for the Seals* Seaweed 13 recreational fishery * assumed was obtained from Kirchner et al. (2000) and for tuna from Armstrong et al. (2004). For the seaweed industry a value of 50 jobs was obtained from Anon. (2005) while Critchley et al. (1991) estimated that the industry supported 250 jobs in 1986. These jobs were converted to jobs/catch value by dividing them by the value of each catch (Table 3). For seals, Hugo (2006) suggests that between 14 and 150 migrant and part-time workers cull seals from August to November.  Model compartments In this section the different functional groups of the model is described including the input data used in the model construction, and where necessary the data used to split groups into adult and juvenile groups. In addition, the spatial distribution of groups are described as input data used for the Ecospace model (see habitat map in Figure 26). 1. Marine mammals  Table 3. Number of jobs per catch value for each of the fleets operational in Namibia Fishery Jobs/value Purse seine 52 Midwater 1 Demersal 9 Longline 96 Crab 42 Lobster 28 Commercial line fishery 14 Recreational line 25 Seals 46 Seaweed 6  Hart and Currie (1960) found that there were not many whales in the Benguela and they attributed this to the cold water, since specifically sperm whales do occur in Saldanha Bay and north of the Benguela but not in the cold upwelling areas. The marine mammals (other than fur seals) off Namibia include southern right whales (Eubalaena australis), humpback and minke whales (Balaenoptera acutorostrata) and small Odontocetes (Roux et al., 2001). Humpback whales (Megaptera novaeangliae) and Bryde’s whales (Balaenoptera edeni) also occur off the west coast of southern Africa (Griffiths et al., 2004). Sightings and surveys of southern right whales indicate that there were at least one adult and calf in 1971 and maybe about 10 adults and one calf in 1999, but there is  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  35  no defendable trend in population size in the data (Roux et al., 2001). Du Pasquier (1990, in Best, 2006) listed 7 whaling grounds on the west coast of southern Africa that were utilized by French whalers in the late eighteenth century, of which Walvis Bay (23°S), Elizabeth Bay (27°S) and Alexander Bay (28°S) are in Namibian waters. From these descriptions and Findlay et al. (1992), it is evident that to some extent marine mammals occur in all areas of the northern Benguela ecosystem. The Odontocetes include dusky (Lagenorhynchus obscurus) and Heaviside’s (Cephalorhynchus heavisidii) dolphins, with a biomass of 1,220 tonnes in the 1980s, although Shannon and JarreTeichmann (1999) quadrupled that biomass to include all other species of whales and dolphins. The bottlenose dolphin (Tursiops truncatus) was found near shore in the vicinity of Walvis Bay and further north, and the southern right whale dolphin (Lissodelphis peronii) is know to occur off Lüderitz (Findlay et al., 1992). The production to biomass ratio (P/B, 1.0 year-1) and production to consumption ratio (P/Q, 7.9%) as well as diet from the 1980s model was used: 1.4% anchovy, 1.4% sardine, 19.3% lanternfish, 6.9% goby, 11.7% small pelagics, 27.5% hake and 31.8% cephalopods (Shannon and Jarre-Teichmann, 1999). The juvenile hake was reduced to 10% and 17.5% added to adult hake in the diet composition. 2. Seals There are 15 breeding and 4 non-breeding Cape fur seal (Arctocephalus pusillus pusillus) colonies along the Namibian coast (De Villiers et al., 1997). According to Mecenero et al. (2006) who wrote on the spatial distribution of the diets of seals, there are two main areas of seal aggregations: Cape Cross (21°47’S and 13°57’E, with 187,000 seals in 2001) and Lüderitz (Van Reenen Bay, Atlas Bay and Wolf Bay, south of Lüderitz, around 27°S and 15°E, with 173,000 seals in 2001). These two areas and the 200 km forage range around them were classified as essential habitat for seals (Mecenero et al., 2006). Hampton (2003) gave numbers of seals extrapolated from the aerial pup sensus, which showed that there were smaller hallouts for seals at Cape Frio and between Sandwich Harbour and Sylvia Hill not shown by Mecenero et al. (2006). Seals are therefore prevalent in the Skeleton Coast Park and the West Coast Recreational Area of the spatial model. South African fur seals were harvested off Namibia since the nineteenth century, with between 75 and 296 tonnes of fish and oil (from shark liver, seals and whales) being exported from Sandwich Harbour between 1863 and 1876 (Kinahan, 1991). Sealing off Namibia was not controlled until 1922, when the Sealing and Fishing Proclamation was passed, and since 1973 sealing was managed under the Sea Birds and Seals Protection Act (David, 1989). The number of pups, bulls and cows harvested in all of southern Africa between 1901 and 1992, the modelled number of pups and total population between 1920 and 2013 and the average weight of pups (age 1=23 kg) and adults (90 kg) were obtained from Butterworth et al. (1995), while the average weight of bulls (150 kg) was obtained from Griffiths et al. (2004).  Catch (tonnes)  From Appendix II in Butterworth et al. (1995) it is estimated that between 1972 and 1993 on average 63% of the fur seal population lived on the Namibian coast. Catches of seal pups were given by David (1989) and by Butterworth et al. (1995) for pups, bulls and cows separately, but these estimates were for the whole population (including South Africa). The same ratio of 63% was 3,000 applied to the catch and the average 2,500 weights for pups, cows and bulls (Butterworth et al., 1995) were used to 2,000 calculate the total catch from 1950 to 1979 (Figure 4). From 1980 to 1990 1,500 the ratio between catches made in South Africa and Namibia obtained 1,000 from Wickens et al. (1991) was used, and since 1990 no seals were caught in 500 South Africa (Griffiths et al., 2004). The catches of pups and bulls in 0 Namibia between 1993 and 2001 were 1950 1960 1970 1980 1990 2000 obtained from Griffiths et al. (2004), and the total number of seals caught Figure 4. Catch (tonnes) of fur seals from 1950 to 2002 in the northern Benguela ecosystem. in 2002 was obtained from Nicols (2004), but he did not indicate if they  36  Northern Benguela Ecosystem, Heymans and Sumaila  were pups or adults. Using the ratio of pups to bulls (14:86) obtained from the catch for 1980-2001, the catch was estimated for 2002. Miller et al. (1996) found that thousands of seals are regularly trapped in trawl nets with 0.5% of the population dying this way (Wickens, 1994 in Miller et al., 1996) and that 66% of the seal bycatch was trapped in midwater research vessels with only 34% trapped by bottom trawlers.  Biomass ('000 t)  Pup numbers ('000)  According to Shannon and Jarre-Teichmann (1999) the biomass was estimated at 51,763 tonnes. Using an average weight of 80 kg for adults with the Butterworth et al. (1995) time series gives a similar average biomass for the 1980s; thus that is used here, giving an increase biomass from 18,000 tonnes in 1950 to 90,000 tonnes in 2005. However, in the 1990s the population declined by more than a third due to lack of food, down from 922,396 seals in 90 450 1992–1993 to 476,074 in 2001– 2002 (Roux, 1999 in Hampton, 2003). Assuming this was a linear 60 300 decline and using a proportion of 78% adults in the population the total population was estimated until 2001 and assumed to be 30 150 constant until 2003 (Figure 5). The estimate of biomass (0.289 calculated from t·km-2) 0 0 Butterworth et al. (1995) was too 1920 1930 1940 1950 1960 1970 1980 1990 2000 high to fit the model and it was reduced to 0.15 t·km-2. Estimated biomass  Observed number of pups  The P/B ratio of 0.94 year-1 was Figure 5. Biomass (*1000 tonnes) estimated and number of pups obtained from Shannon and Jarre(thousands) observed for fur seals in the northern Benguela ecosystem. Teichmann (1999) and according to Balmelli and Wickens (1994) the Q/B for seals is 19.4 year-1. However, the P/B ratio of 0.94 year-1 was much too high and a P/B of 0.08 year-1 was used to fit the model. In addition a biomass accumulation of 0.005 t·km-2·year-1 was included to increase the population to fit the model. Much of the diet of seals is made up of fish; pelagic goby, horse mackerel and juvenile hake are the most important species (David, 1989 in Boyer and Hampton, 2001), but reports of fur seal predation on penguins have increased (Crawford et al., 2001). Between 1974 and 1985 the diet was reported to be 3.6% anchovy, 23.6% horse mackerel, 2.6% sardine, 0.7% lanternfish, 52.9% goby, 4.6% snoek, 3.5% hake and 8.4% cephalopods (David, 1987). The 53% of gobies was divided between gobies and other small pelagics using the ratio of those two species in the diet suggested by Mecenero et al. (2006) giving a breakdown of 6% gobies and 47% small pelagics. The diet given in Mecenero et al. (2006) had a much larger proportion of hake and sardine than that of the earlier years. Mecenero et al. (2005) found that only 0.1% of the frequency of the diet of female fur seals contained bird feathers, specifically those of penguins and cormorants, but reports that Rand (1959) in Mecenero et al. (2005) found 0.8%. Thus we added that percentage to the diet, and at first a value of 1% was used. According to Cury and Shannon (2004) and references therein, seals fed on sardines prior to the sardine decline and have since switched to goby, supplemented by myctophids in the southern part of the northern Benguela and horse mackerel in the central and northern parts. To balance and fit the model the 4.6% snoek in the diet of seals was reduced to 2.6% and 0.01% added to linefish. Adult anchovy in the diet was increased from 3.7% to 50% to fit anchovy; adult sardine was increased to 10.1% and small pelagics were reduced to 3.7% to both these groups. The 23.6% of juvenile horse mackerel was reduced to 10% juveniles and 3.6% added to adult horse mackerel, adult sardine and adult anchovy respectively. Juvenile hake was reduced to 1% and 2.5% added to adult hake. The 8.4% of cephalopods were reduced to 3% and the rest added to sardine. 3. Birds The gannets (Morus capensis) and cormorants (Phalacrocorax capensis) are widespread along the Namibian coast, but the penguins (Spheniscus demersus) tend to be found close to their breeding range and their northern limit is 25°S (Hampton, 2003). In addition to these species, the shy (Thalassarche  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  37  cauta), black-browed (T. melanophris) and yellow-nosed (T. chlororhynchos) albatrosses and whitechinned petrel (Procellaria aequinoctialis) are also found (Ryan et al., 2002), although they do not necessarily breed in Namibia. Cape gannets breed on islands off southern Namibia and Cape cormorants breed on the nearshore islands and guano platforms (Boyer and Hampton, 2001). African penguins also breed on islands off Namibia, and sardine and anchovy used to be their main prey, but now, as with cormorants, pelagic goby is more important as sardine is scarce (Crawford et al., 1985). Cormorants are generally dependent on large, surface-schooling fish with anchovy and sardine being their preferred prey species (Crawford et al., 1985). Other bird species also found in Namibia include great, crowned (Phalacrocorax coronatus) and bank cormorants (P. neglectus), white pelicans (Pelecanus onocrotalus), kelp gulls (Larus dominicanus vetula), gaint petrels (Macronectes giganteus), Pintado petrels (Daption capense), great-winged petrels (Pterodroma macroptera), white-chinned prions (Procellaria aequinoctialis), Cory’s shearwater (Calonectris diomedea), storm petrels (Hydrobates pelagicus), Sandwich terns (Sterna sandvicensis), Damara terns (S. balaenarum) and Arctic terns (S. paradisaea) (Crawford et al., 1991; Best et al., 1997).  Number of adults ('000)  Number of nests ('000)  Four islands off Namibia account 35 120 for 97% of the population of 30 100 African penguins in Namibia, namely Possession, Halifax, 25 80 Ichaboe and Mercury islands 20 (Kemper et al., 2001). Since 60 1990 the breeding population 15 decreased by 3.7% per year, 40 10 which is shown by the active nests on these four islands 20 5 (Kemper et al., 2001). The general trend in penguin 0 0 numbers from 1956 to 2000 is 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 also given by Kemper et al. (2001), showing a decline of Active nests Adults 2.8% per year from the Figure 6. Number of penguin nests and adult penguins (thousands) in approximately 100,000 adults in the northern Benguela ecosystem from 1950 to 2000. the 1950s (Figure 6). The decline between 1950 and 1985 is due to the exploitation of their eggs (Frost et al., 1976 in Kemper et al., 2001) and scarcity of food (Kemper et al., 2001). Similarly, gannets declined to 21% of their mid-1950s population, although cormorants seem to have increased to more than double the 1950s population (Crawford et al., 1991). The changes in these species show that their biomass changed from 1,934 tonnes to 1,272 tonnes between the 1950s and 1980s. Crawford et al. (1991) give estimates of bird biomass of 1,867 tonnes in 150,000 km2 and a P/B of 0.16 year-1 with a Q/B of 120.3 year-1. Using the change in biomass of the penguins, cormorants and gannets between the mid-1950s and 1980s, a biomass of 2,532 tonnes or 0.017 t·km-2 was estimated for the 1950s model if we assume that all other species stayed constant. Seabirds are often caught in longlines, specifically tuna longlines (Ryan et al., 2002). The bycatch rate in the late 1990s on the west coast of South Africa was 0.48 birds killed per 1,000 hooks set (Ryan et al., 2002), while the bycatch rate in the southern hemisphere is on average 0.4 birds per 1,000 hooks (Ryan et al., 2002). In South Africa it was estimated that in the late 1990s between 19,000 and 30,000 birds, of which 70% were albatrosses, were caught annually (Ryan et al., 2002), but no estimate is available for the Namibian longline fishery. Penguins feed on shoaling epipelagic fish, such as anchovy and sardine, and regional trends in their abundance are associated with trends in their prey (Crawford et al., 2001). As sardine have collapsed, the range of penguins in Namibia contracted northwards, with populations at Mercury and Ichaboe islands not being affected by the collapse of sardine as pelagic gobies provided alternative food at these two islands (Crawford et al., 2001). The diet of seabirds was estimated at 0.3% copepods, 4.3% euphausiids, 4.4% cephalopods, 1.1% sardines, 21.4% anchovy, 0.5% horse mackerel, 5.3% mesopelagics, 46.6% gobies, 3.7% other pelagics, 10.3% hake, 0.2% seals and 0.1% seabirds (Shannon and Jarre-Teichmann, 1999). A small fraction of the diet of birds (0.001%) was assumed to be of seal pups on the colony, and the 21.4%  38  Northern Benguela Ecosystem, Heymans and Sumaila  anchovy was increased to 25.3% to fit and the 46.6% of gobies was reduced to 42.9%. The 10% hake was reduced to 1% and 10.8% added to other demersals. As seabirds are a large group that includes both breeding and non-breeding seabirds, this group was assumed to occur in all areas of the spatial model. 4. Sharks  120  100 90 80  Catch rate (kg/hour)  Biomass ('000 t)  100 Three shark species were obtained from the demersal assemblages 70 80 off Namibia in the 1990s, namely 60 the soupfin shark (Galeorhinus 60 50 galeus), African sawtail catshark 40 (Galeus polli) and Izak catshark 40 30 (Holohalaelurus regani) 20 (Hamukuaya et al., 2001). In 20 addition, Macpherson and 10 Gordoa (1992) caught dogfish 0 0 profundorum), (Deania 1980 1985 1990 1995 2000 thornback skate (Raja cf. clavata), blancmange skate (Raja Biomass Catch rate wallacei) and Raja confundens, yellowspotted catshark Figure 7. Biomass estimated (*1000 tonnes) and catch rates in kg/hour (Scyliorhinus capensis) and for sharks and skates in the northern Benguela ecosystem from 1980 to longnose spiny dogfish (Squalus 2000. blainvillei). Finally, Ebert (1996) found that the sevengill shark (Notorynchus cepedianus) was one of the more common species of elasmobranchs caught during angling competitions and they did not travel very far. As the shark group includes pelagic and benthic sharks, rays and skates, it is assumed that sharks occur in all areas of the spatial model.  Kinahan (1991) suggested that shark were already being fished off Namibia for liver oil in the 1800s and were fished out by 1885. The bycatch of sharks was obtained from the Sea Around Us webpage and divided into the demersal, midwater, hand line and other line fisheries in relation to the total catches for those fisheries in this database. The line fisheries were divided into longline and bait boat fisheries in the ratio that tuna were caught by those gears. Biomass estimates for 1983–1990 (Figure 7) were obtained from Macpherson and Gordoa (1992) and were in the same range as the 67,180 tonnes reported by Shannon and Jarre-Teichmann (1999), with an average of 73,608 tonnes. This could be used as a first estimate for biomass, although the 1950 biomass might be higher due to reductions by the hake fishery and other demersal fisheries. Bianchi et al. (2001) gave estimates of shark catch rates for 1990–2000 and Hamukuaya et al. (2001) found that in the early 1990s, Chondrichthyans on average made up 3.6% of the weight of the taxa caught in the demersal surveys, which could indicate that their biomass could be about 3.6% of that of all demersal species, including hake, monkfish, cephalopods, crabs and other demersal fish. Using 3.6% of the biomass of hake, monkfish, cephalopods and crabs only estimates a biomass of approximately 35,000 tonnes for the 1980s; thus a starting biomass similar to the average obtained from Macpherson and Gordoa (1992) was used, but the time series of biomasses were estimated from the biomasses of all other species as that is a relative estimate (Figure 7). The P/B and P/Q and diet estimates used for the 1980s were also employed here: 0.5 year-1 and 25% respectively, and the diet consisted of 9.3% anchovy, 1.2% sardine, 1.2% horse mackerel, 11% small pelagics, 0.4% large pelagics (split into 0.15% snoek, 0.15% linefish, 0.1% tuna), 10% cephalopods, 41.5% demersals (which was divided into 10% monkfish and 31.5% other demersals), 0.4% hake, 10.4% macrobenthos (including crabs), 0.2% seals and 14.4% cannibalism (Shannon and Jarre-Teichmann, 1999). To fit the model the 0.2% seals had to be reduced to 0.1% and 0.1% was added to other marine mammals. Similarly, the demersals were reallocated to 5% monkfish and 36.5% other demersals. The macrobenthos was split between macrobenthos (5.4%) and crabs (5%).  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  39  5. Tunas and large pelagics  25 20 15 10  5 0 1956  1961  1966  1971  Bait boats  1976  1981  Longline  1986  1991  Purse seine  1996  2001  Other  Figure 8. Tuna catch by fleet (*1000 tonnes) for the northern Benguela ecosystem. 200  Biomass ('000 t)  Tuna has been fished since 1956 and the catches were obtained from the CATDIS dataset of ICCAT for all species including the marlins, sailfish and swordfish (www.iccat.es/ downloads.htm) for the area 15°S–30°S and 10°E–15°E. The ICCAT data was broken down by fleets: longline, purse seine, bait boat and other gear (Figure 8). The catches of other large pelagics were obtained from the Sea Around Us database and added to the ICCAT data for this group. Biomass estimates for albacore in the South Atlantic were obtained from Punt et al. (1995) and only used as a relative biomass for this group (Figure 9).  Catch ('000 t)  Albacore (Thunnus alalunga), yellowfin (T. albacares) and bigeye tuna (T. obesus) are all important tuna species in the catches off Namibia (Shannon and Jarre-Teichmann, 1999). Other species also recorded in the catches include northern bluefin tuna (T. thynnus), skipjack (Katsuwonus pelamis), Atlantic sailfish (Istiophorus albicans), Atlantic blue marlin (Makaira nigricans), Atlantic white marlin (Tetrapturus albidus) and swordfish (Xiphias gladius). All of these species are large pelagics that roam the world oceans from the shelf to the upper pelagic over the deep ocean; thus the tuna group was assumed to occur in the shelf, slope and deep ocean 35 sections north and south of the Walvis Ridge in the spatial 30 model.  150  100  The natural mortality for tuna was assumed to be 0.2 year-1, which 50 was used for the P/B ratio and the P/Q was assumed to be 5%. The diet consisted of 20% euphausiids, 0 7% myctophids, 2% gobies, 15% 1956 1961 1966 1971 1976 1981 1986 1991 1996 anchovy, 15% sardine, 16% small pelagics, 5% horse mackerel, 15% Figure 9. Biomass estimates (*1000 tonnes) for albacore tuna in the cephalopods and 5% hake southeast Atlantic used as a proxy for tuna biomass in the northern (Shannon and Jarre-Teichmann, Benguela ecosystem. 1999). No estimate of biomass was available, but to fit the model a biomass of 1 t·km-2 had to be assumed and the P/B increased to 0.25 year-1. To fit the model the diet of tuna was changed to include 0.6% juvenile hake and 2.4% adult hake and 37.5% import had to be imported into the model as this group roams beyond the confines of the northern Benguela. 6. Snoek Snoek (Thyrsites atun) is a pelagic predator that has been recorded from northern Angola to the east coast of South Africa (Griffiths, 2003). The distribution of snoek in the northern Benguela was described by Griffiths (2003) to span from south of Lüderitz to around Cape Frio and north, mainly on the continental shelf. It occurs in a depth range from 0 to 550 m (Froese, 2000); thus it was assumed to occur in the surf zone, on the shelf and on the slope of the whole spatial model of the northern Benguela ecosystem.  40  Northern Benguela Ecosystem, Heymans and Sumaila  Snoek has been an important commercial species since the 1800s, caught first with hand lines and later with trawlers (Griffiths, 2003), with snoek caught as a bycatch in the midwater trawl fishery (Bianchi et al., 1993). Between 1972 and 1980, 90% of the overall catch of snoek in the southeast Atlantic was taken in the area that sustains the Namibian and South African purse seine fisheries, with very little being caught in ICSEAF division 1.5 around the Lüderitz upwelling cell (Crawford and De Villiers, 1985). The general consensus was that there is only one stock of snoek on the west coast of southern Africa, but Griffiths (2003) suggested that the snoek found off Namibia and South Africa are separate sub-populations, although there could be extensive interaction between the populations. 8,000 Catch (tonnes)  No estimate of snoek biomass was available, but the total catch time series of snoek made with hand line and midwater trawls was obtained from Griffiths et al. (2004). Prior to 1972 snoek caught by the trawl fishery was discarded, and therefore there is no estimate of catches by the trawl fishery before 1972 (Griffiths et al., 2004). From 2001 to 2003 estimates of snoek catches were obtained from the Sea Around Us database and were assumed to be made mostly by the hand line fishery (Figure 10). The P/B and P/Q ratios of 0.25 year-1 and 10% obtained from Shannon and Jarre-Teichmann (1999) were used.  6,000 4,000 2,000 0 1950 1956 1962 1968 1974 1980 1986 1992 1998 Linefish  Snoek  Snoek bycatch midwater fleet  Figure 10: Catches of snoek and linefish (tonnes) made by the commercial line fishery and snoek bycatch by the midwater fleet.  Griffiths (2002) gives estimates of diet for snoek on the west coast of South Africa that include 12.6% macrobenthos, 3.5% euphausiids, 0.7% cephalopods, 12.3% anchovy, 17.5% sardine, 7.2% small pelagics, 16.1% mesopelagics, 5.4% horse mackerel, 9.8% demersals, 0.5% gobies and 14.3% hake. The 12.5% anchovy in the diet was reduced to 5% and the 16.1% mesopelagics were reduced to 5% with the difference added to mesozooplankton. The allocation for juvenile horse mackerel (5.4%) was reduced to 1% and 4.4% added to adult horse mackerel to balance juveniles. Juvenile hake was reduced to 0.3%, with 1% added to adult hake and the rest to macrozooplankton. The 9.8% demersals were reduced to 1% to balance that group. The 12.6% macrobenthos was split between 9% macrobenthos, 1% crabs and 2.6% lobster. 7. Linefish (steenbras and kob) The two main species of linefish caught by the Namibian surf-and-rock recreational fishery and the commercial line fishery are steenbras (Lithognathus aureti) and kob (Argyrosomus spp.), but linefish also include the strepie (Sarpa salpa), breams (Pagrus spp.) and blacktail (Diplodus spp). Linefish are caught mainly in the West Coast Recreational Area, although there is also limited access to linefish in the Skeleton Coast Park, Namib Naukluft Park, Sperrgebiet and the surf zones north and south of these areas. Linefish occur mostly in the surf zone and on the shelf, but also over the slope areas on occasion (Froese and Pauly, 2000). Both kob and steenbras have been caught since 1964 by the commercial fishery (Venter, 1988b), but also before that according to the Sea Around Us database. The catch of kob by the commercial fishery from 1964 to 2000 was obtained from Griffiths et al. (2004) and the catch of steenbras between 1964 and 1972 was obtained from Venter (1988b) and from 1973 to 1999 from Holtzhausen (1999). The catches of both kob and steenbras from 1950 to 1963, the catch of steenbras from 2000 to 2003 and kob from 2001 to 2003 and all catches of blacktail, breams and strepie were obtained from the Sea Around Us database (Figure 10). However, these catches were only for the commercial fishery as no recreational catches were available. Holtzhausen and Kirchner (2004) also give catches of kob by ski boat, line boat and anglers and catches of steenbras by anglers for 1995–2000. Holtzhausen and Kirchner (2004) give the best estimate of silver kob and steenbras biomass in 2000 as 7,175 tonnes and 2,006 tonnes respectively and found that they were respectively 40% and 53% depleted  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  41  from their unexploited states, giving a combined biomass in the 1950s of 21,722 tonnes. This value was used as a minimum biomass for this group and Holtzhausen et al. (2001) estimated a combined biomass of 9,872 tonnes for both species in 1995. The natural mortality of 0.23 for steenbras and 0.15 for kob was obtained from Holtzhausen et al. (2001), while Shin et al. (2004) estimated an M of 0.23 for kob, thus a P/B of 0.2 was assumed and a P/Q of 10% was taken from Shannon and Jarre-Teichmann (1999). Kob feeds on zooplankton (6.8%), anchovy (51%), sardine (0.6%), horse mackerel (3.3%), cephalopods (6.1%), macrobenthos (8.8%), hake (8.2%), cannibalism (0.4%), demersals (13.7%) and small pelagics (1.1%), and this diet was used for all linefish. The 8% juvenile hake was reduced to 5.5% and the rest added to macrozooplankton and the 13.7% demersals were reduced to 4.1% and the rest added to detritus to balance and fit the model. The 6.1% cephalopods was reduced to 3.3% and the rest added to macrozooplankton. 8. Juvenile anchovy Crawford et al. (1989) found that juvenile anchovy aggregated north of about 24°S to about Henties Bay (22°S), and Mecenero et al. (2006) found that the Lüderitz area had smaller anchovies than around Cape Cross. Thus, juvenile anchovies were assumed to occur in the West Coast Recreational Area, Namib Naukluft Park, Sperrgebiet and on the southern shelf. Anchovy was split in adults and juveniles at age 4 months, with a von Bertallanffy K value of 0.58 (Fishbase, Froese and Pauly, 2000, www.fishbase.org) and a Wmat/W∞ of 0.1. This weight at maturity was obtained from Fishbase (weight at L∞ of 16.6 cm = 18.3 g), the weight at age 0 from le Clus et al. (1988) and the length-weight relationship in Fishbase. Juvenile anchovy consume only phytoplankton. 9. Adult anchovy Adult anchovy occur mainly in the north (Hewitson and Cruickshank, 1993), and Mecenero et al. (2006) found that larger anchovies were present in the scat of fur seals at Cape Cross. Le Clus et al. (1988) give the catches of anchovy in relation to their CPUE by 2° blocks from 17°S to 30°S for 1972–1987. From this paper, it seems that the main catches of anchovy were made between 22°S and 25°S, but the distribution of catches was more spread out than that of sardine, and the largest catches were between 23°S and 24°S (le Clus et al., 1988). Thus, their non-preferred habitat includes the area from the Benguela-Angola front to Henties Bay (22°S) and from Sylvia Hill (25°S) to the Orange River. 800 Catch and biomass ('000 t)  According to Hampton (2003) anchovy was not caught much in Namibia before 1966, but after the collapse of the sardine fishery in the 1970s the catches increased substantially. Anchovy (Engraulis japonicus) only appears in the catch statistics from 1964 onwards, and catches of anchovy (Figure 11) from 1964 to 2000 were obtained from Willemse (2002), while the catches from 2001 to 2003 were obtained from the Sea Around Us database.  600 400 200 0 1963  1968  1973  1978  1983  Catch  1988  1993  1998  2003  Biomass  Figure 11. Catch and biomass (*1000 tonnes) of anchovy in the northern Benguela ecosystem.  A VPA was done on anchovy from 1972 to 1984 by le Clus (1985) in Crawford et al. (1987), but the data were only used for 1972–1979 (Figure 11). The data from 1980 to 1989 came from Hewitson and Cruickshank (1993) and from 1990 to 1995 were obtained from the Ministry of Fisheries and Marine Resources (Heymans, 1997). Hampton (2003) suggests that the biomass in 1996 and 1997 was around 20,000 tonnes and this is used here; however, after 1997 the survey design changed and no estimate of anchovy biomass is available. Even though no estimate of anchovy biomass was available an estimate of 5.7 t·km-2 was used as it fit the model best.  42  Northern Benguela Ecosystem, Heymans and Sumaila  Adult anchovy P/B and Q/B ratios of 1.16 year-1 and 11.7% respectively were obtained from Shannon and Jarre-Teichmann (1999). The P/B for juveniles was estimated at 10 year-1. To fit the model a P/B ratio of 0.85 year-1 was used for adult anchovy. Adult anchovy diet consists of 33% phytoplankton, 4% microzooplankton (detritus), 31% copepods and 32% euphausiids (Shannon and Jarre-Teichmann, 1999) and a small percentage (0.1%) of their diet was assumed to consist of juvenile sardine and juvenile gobies and 0.01% of their diet of juvenile jellyfish. 10. Juvenile sardine Roy et al. (1992) suggest that the relationship between sardine year class strength and sea surface temperature is positive in the southern Benguela but negative in the northern Benguela. While looking at the scat samples of fur seals, Mecenero et al. (2006) found that both the Cape Cross and Lüderitz areas had large numbers of sardine. Crawford et al. (1989) showed that juvenile sardine aggregate north of about 24°S to about Henties Bay (22°S). However, there was a large area of spawning for sardine from 17°S to 21°S (le Clus, 1990), and O’Toole (1977) in Maurihungirire (2004) found that Namibian sardine spawn within 60 km of the coast between 21°S and the confluence of the Benguela and Angola currents in late summer or autumn in water above 19°C. Thus, in this model adult and juvenile sardine were assumed to occur in all areas of the spatial model except the deep ocean and slope areas. Sardine was split into juveniles (12 months) and adults using the ratio of weight at age 1 (91 g) to weight at infinity (342 g) obtained from le Clus et al. (1988) and Fishbase respectively and a K value of 0.43 also obtained from Fishbase. For juveniles the P/B of 10 year-1 was assumed. Juvenile sardine consume only phytoplankton. 11. Adult sardine Le Clus et al. (1988) give the catches of sardine in relation to their CPUE by 2° blocks from 17°S to 30°S for 1972–1987, which could be used to verify the estimates of catches by year for sardine. From Le Clus et al. (1988) it seems that the main catches of sardine were made between 17°S and 25°S, with the highest catches between 22°S and 23°S (le Clus et al., 1988). Hewitson and Cruickshank (1993) also suggested that sardine is mainly found in the north. However, Mecenero et al. (2006) found that both the Cape Cross and Lüderitz areas had large numbers of sardine; thus adult sardine were assumed to occur in all areas of the spatial model except the deep ocean and the slope areas.  The diet of sardines consists of phytoplankton (56%), microzooplankton (detritus, 8%), copepods and euphausiids (18% each) (King and Macleod,  12  1.6  9  1.2  6  0.8  3  0.4  0 1950  0 1960  1970 Biomass  1980  1990  2000  Catch  Figure 12. Catch and biomass (million tonnes) of sardine in the northern Benguela ecosystem.  Catch (million t)  Shannon and Jarre-Teichmann (1999) used a P/Q of 10% which gives a Q/B ratio of 5.0 year-1. Natural mortality was estimated at 0.59 year-1 in the 1960s (Newman, 1970 in Fossen et al., 2001), and (Butterworth, 1983 in Fossen et al., 2001) used 0.5 year-1 for his VPA, which is what was used here for adults. Fossen et al. (2001) estimated a value that varies between 0.77 and 2.38.  Biomass (million t)  Catches of sardine from 1950 to 2000 (Figure 12) were obtained from Willemse (2002), for 2001 to 2002 from Nicols (2004) and for 2003 from the Sea Around Us database. Biomass estimates for sardine were obtained from Schwartzlose et al. (1999), le Clus et al. (1988) and Thomas (1986) for 1952 to 1983 and from 1984 to 1999 from Kreiner et al. (2001), while the biomass for 2000 and 2001 was obtained from Hampton (2003).  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  43  1976 in Shannon and Jarre-Teichmann, 1999). A small percentage (0.1%) of their diet was assumed to consist of juvenile anchovy and juvenile gobies and 0.01% of their diet of juvenile jellyfish. 12. Juvenile gobies Mecenero et al. (2006) found from scat samples of fur seals that the Cape Cross area has smaller gobies while the Lüderitz area has larger gobies, and Crawford et al. (1989) showed that juvenile gobies aggregate around Walvis Bay (17°S–20°S). However, in the spatial model of this ecosystem juvenile gobies were assumed to be distributed throughout the surf zone and shelf areas of the ecosystem but not on the slope. Migration of juveniles from the north to the south and adults from the south to the north were instituted. Gobies were split into juvenile and adults at age 4 months, with a K of 0.4 and a Wmat/W∞ of 0.2 (Fishbase) and a P/B for juveniles of 7 year-1. They only consume phytoplankton. 13. Adult gobies  Catch (tonnes)  Catch rate (kg/hr)  Gobies are found up to 60 km offshore (Cruickshank et al., 1980) and small gobies are more pelagic than larger individuals, which are mainly bottom dwellers and found further offshore (Crawford et al., 1985), although seals do seem to feed on larger gobies near the surface at night (Mecenero et al., 2006). Hamukuaya et al. (2001) showed that gobies were mostly found on the central shelf (20–200 m) and sometimes on the southern 1400 100 shelf (south of Elizabeth Bay at 27.5°S). Pelagic gobies were 1200 more important in the south, 80 and the highest concentration 1000 of gobies was found between 60 800 Lüderitz and Walvis Bay (Hewitson and Cruickshank, 600 40 1993). In the spatial model 400 adult gobies were assumed to 20 occur in the surf zone, shelf 200 and slope areas of both the northern and southern part of 0 0 the ecosystem but not in the 1980 1985 1990 1995 2000 deep ocean. Catch  Catch rate  According to references in Figure 13. Catch (tonnes) and catch rate (kg/hour) for gobies in the northern Benguela ecosystem. Melo and le Clus (2005), biomass estimates in the northern Benguela range between 0.6 and 1.45 million tonnes (Shannon and Jarre-Teichmann, 1999), but Hewitson and Cruickshank (1993) estimated 600,000 tonnes between 1978 and 1983, so that was used as a starting biomass and the P/B and P/Q was 0.9 year-1 and 10% respectively (Shannon and JarreTeichmann, 1999). However, to fit the model the P/B and Q/B ratios for adult gobies were assumed to be 1.1 year-1 and 15 year-1 respectively. A biomass for adult gobies of 6.3 t·km-2 was used to fit gobies. Catch rates of gobies in the 1990s obtained from Bianchi et al. (2001) were used as a relative estimate biomass time series (Figure 13). Catches of gobies from 1983 to 2000 were obtained from Willemse (2002). The diet of gobies consists of phytoplankton (93%), 1% copepods and 6% euphausiids (Crawford et al., 1985). A small percentage (0.1%) of their diet was assumed to consist of juvenile anchovy and juvenile sardine and 0.01% of their diet of juvenile jellyfish. 14. Other small pelagics Other small pelagics in this model included chub mackerel (Scomber japonicus) and round herring (Etrumeus whiteheadi) which have been caught since 1971 (Willemse, 2002). In addition, the Sea Around Us database shows that Sardinella aurita, other Sardinellas and other Carangidae, Sparidae and Engraulidae were also caught, and the total catch for this group was then obtained from the Sea Around  44  Northern Benguela Ecosystem, Heymans and Sumaila  Us database (Figure 14). Also included in this group are the flying fish (Exocoetidae) and sauries (Scomberesocidae). The group was assumed to occur in all areas of the spatial model. Very little is known about these species, but their catches were obtained from Willemse (2002), and their P/B and P/Q ratios were estimated at 0.6–0.9 year-1 and 10% respectively (Shannon and JarreTeichmann, 1999). No estimate of biomass was available and it was estimated using an ecotrophic efficiency of 95% for this group. Using the average diet of other small pelagics in chub mackerel and their biomass estimates obtained from Shannon and Jarre-Teichmann (1999), the average diet for this group was estimated at 55.8% copepods, 35% euphausiids, 2.1% jellyfish, 1.4% benthos, 0.5% anchovy, 0.3% cannibalism and 5% mesopelagics. The mesopelagics were reduced to 1% and the 4% added to mesozooplankton (copepods). 15. Mesopelagics  300  4 Mesopelagics ('000 t)  Pelagics ('000 t)  The most important 250 3 mesopelagics in the northern 200 Benguela are the lanternfish (Lampanyctodes hectoris) 150 2 and the lightfish (Maurolicus muelleri) (Hewitson and 100 Cruickshank, 1993). 1 Lanternfish prefer water 50 between 300 m and 1000 m 0 0 deep and migrate vertically (Cruickshank, 1983). They 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 live along the edge of the continental shelf but are Small pelagics Mesopelagics concentrated from Walvis Figure 14. Catch (*1000 tonnes) for small pelagics and mesopelagics in the Bay to the Orange River northern Benguela ecosystem. about 30 km from the shore (Cruickshank, 1983). They are found mainly in the north, but mostly further offshore (Mecenero et al., 2006); thus in the spatial model they were assumed to occur on the shelf, slope and over the deep ocean in both the northern and southern part of the ecosystem. Lanternfish have been recorded in the catches given by the Sea Around Us database from 1970 to 2003 (Figure 14). The lower limit for their biomass was 800,000 tonnes of lanternfish (Hewitson and Cruickshank, 1993) and 10,000 tonnes of lightfish (Shannon and Jarre-Teichmann, 1999) for the northern Benguela and their P/B and P/Q ratios were estimated at 1.23 year-1 and 10% respectively (Hewitson and Cruickshank, 1993). However, for this model of the 1950s the biomass of mesopelagics was estimated by using an ecotrophic efficiency of 95%. Their diet was assumed to be 40% copepods and 60% euphausiids (Hewitson and Cruickshank, 1993). They are caught as bycatch with an estimated catch of 1000 tonnes·year-1 (Shannon and Jarre-Teichmann, 1999) but no time series of catches were available. 16. Juvenile horse mackerel Juvenile horse mackerel are important in the north, from 20°S to the Angolan border (Axelsen et al., 2004). They are found from south of Conception Bay (around 25°S) all the way to the Angola-Benguela front at about 15°S (Axelsen et al., 2004). This was confirmed by Crawford et al. (1989), who showed that juvenile horse mackerel aggregated close to Cape Frio. Mecenero et al. (2006) also found large numbers of horse mackerel in the scat samples of fur seals in the Cape Cross area, while the Lüderitz area had not much horse mackerel. According to Boyer and Hampton (2001) juveniles live inshore of the 100 m isobath and mature fish move offshore, while Axelsen et al. (2004) suggest that they occur inshore from the 200 m depth contour. Thus, in the spatial model juvenile horse mackerel was assumed to occur in the shelf and surf zone areas of the northern part of the system, as well as in the Skeleton Coast Park, West Coast Recreational Area and the Namib Naukluft Park, but not further south. Both juveniles and adults migrate along the shore and across the shelf break, but juveniles are mostly found closer inshore and adults offshore (Maurihungirire, 2004). Thus in the model juveniles were assumed to migrate south in the inshore waters and adults were assumed to migrate north in the offshore waters.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  45  Horse mackerel was divided into adult and juveniles at age 2, with L∞ of 51 cm, M of 0.4 and K of 0.13 or 0.15 (Anon., 2001). Krakstad and Kanandjembo (2001) give a range of values for L∞ between 41.1 and 62.6, a range of K from 0.1 to 0.25, and a range of M from 0.22 to 0.53 obtained from the literature. The values obtained by Kampowski and Slosarczyk (1976) in Krakstad and Kannadjembo (2001) of L∞=47 cm, K=0.25 and M=0.25–0.5 were used to balance the model of the 1950s. The Wmat/W∞ ratio (0.066) was estimated from a Wmat of 92 g for 2-year-old fish (Klingelhoeffer, 2006) and W∞ of 1,356 g at L∞ of 55 cm calculated from the weight-length relationship of W=0.0078 * L3.011 (Fishbase). To fit the model it was necessary to use a P/B of 1.7 year-1 for juveniles. Up to 2 years of age they feed near the surface and are zooplanktivorous with their diet consisting mainly of copepods, similar to that of sardine and anchovy (Venter, 1976 in Boyer and Hampton, 2001). Andronov (1983) suggested that even in 1963 most of the diet consisted of copepods. 17. Adult horse mackerel Similar to juvenile horse mackerel, adults are found from about 25°S to the Angola-Benguela front, but they are mainly found offshore of the 200 m depth contour (Axelsen et al., 2004). Hampton (2003) showed that adult horse mackerel are very prevalent offshore between 18°S and 21°S but, more importantly, inshore north of 18°S and between 21°S and 24°S. In the demersal assemblages, adult horse mackerel is found in the shelf areas, moving north and south over different years, with the main concentration south from the Kunene River (17.5°) to around Elizabeth Bay (27°S) but also occurring south to the Orange River in some years (Hamukuaya et al., 2001). Horse mackerel spawn in warmer water west of the shelf break, and the nursery areas are adjacent to the spawning areas but closer to the shore (Maurihungirire, 2004). The distribution of horse mackerel in the spatial model included shelf and surf zone areas in both the northern and southern parts of the system, including the Skeleton Coast Park, the West Coast Recreational Area, the Namib Naukluft Park and the Sperrgebiet. The adults are found further offshore, and the migration of juveniles from north to south and of adults from south to north was included.  Biomass ('000 t)  Biomass estimates of horse mackerel from 1966 to 2000 were obtained from Klingelhoeffer (2006) for both B0+ and B2+ and the difference was used as the biomass for juvenile horse mackerel. There were also VPA estimates for 1966–1985 from Crawford et al. (1987), for 1991–1993 from acoustic estimates obtained from the Ministry of Fisheries and Marine Resources, for 1994–2000 from Bauleth-D’Almeida et al. (2001), and, finally, Axelsen et al. (2004) give acoustic estimates of horse mackerel biomass from 1989 to 2003. However, the data obtained from Klingelhoeffer (2006) were 3000 used and interpolated from 1987 to 1991, and for 2001 the estimate was obtained from Hampton 2000 (2003) and divided into adults and juveniles by using an average of 72% adult weight in the population as given since 1994 (Figure 15). In 1000 addition, the fishing mortality from 1966 to 1986 and from 1991 to 2000 were also obtained from Klingelhoeffer (2006). There was a 0 reduction in the mean age from 6.5 1966 1971 1976 1981 1986 1991 1996 2001 years to 1.9 years from 1968 to 1971 (Anon., 2001) after the start of the Juvenile biomass Adult biomass fishery. Figure 15. Biomass (*1000 tonnes) of adult and juvenile horse  Catches of adult horse mackerel mackerel in the northern Benguela ecosystem. (Figure 16) made by the midwater trawl fishery from 1955 to 1960 were obtained from Willemse (2002), while catches from 1961 to 2004 of the midwater trawl (adult) fishery and from 1970 to 2004 of the purse seine (juvenile) fishery were obtained from Klingelhoeffer (2006). According to Anon. (2001), the large Ukrainian fleet targeted horse mackerel since 1963. In addition, horse mackerel is also caught as a bycatch by the hake demersal fishery, and the data of these catches between 1980 and 2004 were obtained from Klingelhoeffer (2006).  46  700 600  Catch ('000 t)  As horse mackerel was probably always a bycatch of the demersal fleet, the average of the percentage of horse mackerel to hake in the 1980s was used (0.13%) with the catch of hake from 1964 to 1979 to calculate the bycatch for that time (Figure 16). For 1950–1970 the catch by the purse seine fleet was estimated as the difference between the catch obtained from the Sea Around Us database and the total catch of the midwater trawlers and demersal trawlers.  Northern Benguela Ecosystem, Heymans and Sumaila  500 400 300 200 100 0 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 Purse seine  Midwater  Demersal  The biomass of adult horse Figure 16. Catch (*1000 tonnes) of horse mackerel by the purse seine mackerel needed to be 7 t·km-2 to (juveniles), demersal and midwater fleets (adults). fit the model and it was necessary to use a P/B of 0.4 year-1 for adults. P/B and P/Q ratios for adult horse mackerel of 0.52 year-1 and 10% respectively were used by Shannon and Jarre-Teichmann (1999), which gives a Q/B ratio of 5.2 year-1 for adults. For juveniles a Q/B of 35 year-1 was used at first based on the estimates made by Shannon and Jarre-Teichmann (1999) based on work by Andronov (1985). The diet of adult horse mackerel was 95% euphausiids off Namibia (Konchina, 1986; Boyer and Hampton, 2001), but they also feed on copepods and to a lesser extent on lantern fish (Andronov, 1983). Shannon and Jarre-Teichmann (1999) estimated the diet of adult horse mackerel to include 18% copepods, 78% euphausiids, 0.9% gobies, 0.1% cephalopods and 2% lanternfish, while Andronov (1983) estimates 43% each for copepods and euphausiids, 10% lanternfish, 0.5 cephalopods and 3.1% gobies. However, Andronov (1983) estimated percent abundance not percent mass, thus this model used the estimates of Shannon and Jarre-Teichmann (1999). 18. Juvenile hake Two species of hake are prevalent off Namibia, namely Cape hake, Merluccius capensis and M. paradoxus. Burmeister (2005) indicates that M. paradoxus might be a single stock in South Africa and Namibia and not two separate stocks, so it might be difficult to model their ontogenetic split in the northern Benguela alone, as good recruitment would not necessarily have an effect on the Namibian M. paradoxus stock. M. paradoxus also does not spawn off Namibia and recruitment is only found in southern Namibia (Gordoa et al. 1995 in Maurihungirire, 2004). Crawford et al. (1989) showed that juvenile hake aggregate north of about 24°S to about Henties Bay (22°S), but they are also found from about the Kunene (17°S) to Cape Cross (22°S) and from south of 24°S to Lüderitz. Mecenero et al. (2006) also found smaller hake in the scats of fur seals around Cape Cross, while the larger hake seemed to occur in the Lüderitz area. In the spatial model juvenile hake was assumed to occur on the slope and shelf of both the northern and southern areas of the model as they were not totally excluded from these areas by any references. Shin et al. (2004) give estimates of K (0.046 year-1) and Zo for age 0 fish (8 year-1), while Botha (1971) in Burmeister (2005) indicates that hake adopt a demersal habitat at a length of 15–20cm at roughly an age of two years. The Zo for age 0 fish (8 year-1) was used as the P/B for juveniles. However, this ratio is too high for juveniles up to 24 months, and it was reduced to 1.05 year-1 to be more realistic. Hake growth parameters of a=0.005 and b=3.11 were obtained from Fishbase, and with the L∞ of 116 cm total length, a W∞ of 13,165 g was estimated. Similarly, an average length at maturity of 40 cm was obtained from Fishbase and a Wmat of 480 g was estimated, giving a Wmat/ W∞ ratio of 0.036.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  47  Boyer and Hampton (2001) suggest that hake are opportunistic feeders, with young M. capensis and M. paradoxus feeding predominantly on planktonic crustaceans such as euphausiids, pelagic gobies and lanternfish. Huse et al. (1998) give estimates of M. capensis (which were mostly larger than 50 cm) and M. paradoxus (mostly smaller than 50 cm) diets, which were used as the adult and juvenile diets in conjunction with Pillar and Barange (1997) and Andronov (1983) to disaggregate the mixed contents and miscellaneous groups given in Huse et al. (1998). For juveniles the diet was broken down into: 81.5% euphausiids, 9.8% mesopelagics and 1.4% each for cephalopods, macrobenthos, small pelagics, anchovy and sardine. Roel and Macpherson (1988) found that the diet of hake is determined by the availability of prey and therefore this will change with balancing. In addition, this diet was based on percentage frequency of occurrence and not percentage weight. The juvenile diet was changed to balance and fit the model by reducing the mesopelagics in their diet from 9.8% to 1% and adding the rest to mesozooplankton. 19. Adult hake  1.5  900  1.0  600  0.5  300  0.0 1950  Catch ('000 t)  Shin et al. (2004) give estimates of M (0.5 for M. capensis and 0.4 for M. paradoxus) for adults and the average (0.45 year-1) was then used as the P/B for adults. The Q/B ratio for juvenile hake was estimated at between 5.9 and 17.2 year-1 while Jarre-Teichmann et al. (1998) used 2.4 year-1 for hake aged 3 years and older and 8.0 year-1 for younger hake. These values are used for this model but Punt et al. (1992) estimate a value of ~3% per day (10.9 per year) and Jarre-Teichmann et al. (1998) used 11.6 year-1.  VPA biomass (million t)  Hampton (2003) showed that Cape hake (M. capensis) occurred in large numbers between 200 m and 500 m depth from the Kunene to Cape Frio and from about 23°S to north of Lüderitz, and south of Lüderitz to the Orange River, while deepwater hake (M. paradoxus) were more prevalent offshore (500– 1000 m depth) between Cape Frio and 28°S with some lesser aggregations at the Orange River and Easter Point (25.5°S). This is similar to what was found by Maurihungirire (2004), that M. paradoxus seems to dominate where the continental slope is less steep and the shelf is narrow, while M. capensis is abundant on wider continental shelves. Punt et al. (1992) showed that Cape hake occur between 20°S and 30°S but that deepwater hake only occur north to about 25°S from South Africa. Similarly, Abelló et al. (1988) showed that shallow water hake occurred in larger numbers around Walvis Bay and the Orange River, while deepwater hake occurred in large numbers offshore from 26°S to the Orange River, although their study did not go further north than 23°S. Mecenero et al. (2006) found larger hake in the scats off Lüderitz than off Cape Cross in the north. In the spatial model adult hake was assumed to occur on the slope and in the deep waters of both the northern and the southern part of the model.  0 1960  1970  Midwater  1980  Demersal  1990  2000  VPA biomass  The catch statistics for hake have Figure 17. Hake biomass (million tonnes) estimated from VPA and always been reported for the two catch (*1000 tonnes) made by the midwater and demersal fleets. species combined (Figure 17). Klingelhoeffer (2006) suggests that the demersal trawl fishery started in the 1950s and therefore catches from 1950 to 1963 were obtained from the Sea Around Us database. Similarly the catches from 2001 to 2003 were obtained from the Sea Around Us database, while from 1964 to 2000 the catches of hake were obtained from Willemse (2002). Hake is also caught as bycatch in the midwater trawl horse mackerel fishery, and according to Klingelhoeffer (2006), the bycatch ranged from 0.1%–2% for the Polish fleet during 1980–1984 to 18% for the Soviet fleet in 1990. Thus using an average of 7% of the landings of the midwater trawl fishery being hake gave an estimate of hake bycatch by the midwater horse mackerel fleet. As hake adopts a demersal habitat at age 2 (Botha, 1971 in Burmeister, 2005), this catch is assumed to be of juvenile hake. Hampton  48  Northern Benguela Ecosystem, Heymans and Sumaila  (2003) found that by 2001 the bycatch was approximately 2% of the total midwater catch. Hake is also caught by the longline fleet, and in 1998 16,500 tonnes of hake were landed by this fleet (Hampton, 2003). CPUE estimates for hake in the ICSEAF divisions were obtained from Butterworth and Rademeyer (2005), who indicated that the values from 1965 to 1980 were more reliable than those from 1981 to 1988 when there was misreporting of catches. Hake biomass for both species were estimated by VPA for ICSEAF divisions 1.3+1.4 from 1968 to 1984 and for division 1.5 from 1965 to 1984 (Leslie, 1986 in Crawford et al., 1987), while Van der Westhuizen (2001) gives estimates of biomass from 1990 to 2000 (Figure 17). Hampton (2003) estimates that the 2001 spawning biomass for both species of hake combined was approximately 1.1 million tonnes (similar to that obtained for 2000 by Van der Westhuizen (2001)). For 1985–1989 estimates were obtained from Macpherson and Gordoa (1992) and were prorated to be in the same order of magnitude by using the value for 1984 obtained from VPA and the surveys. Butterworth and Geromont (2001) gave a stock assessment trajectory for hake 2+biomass if the estimates of the Dr. Fridtjof Nansen were assumed to be relative estimates of biomass. This trajectory from 1965 to 1999 was 5 times more abundant than the previous trajectory but did have a similar trend; thus it was used as a relative estimate of biomass. Surplus production models were used to estimate the biomass during the ICSEAF management phase, but there are concerns about the reliability of these assessments (Hampton, 2003). However, we used the values given by Butterworth and Geromont (2001) for the relative biomass for 2+, which gives a biomass of 24 t·km-2 in 1966, which we used for 1956 (Figure 17). Punt et al. (1992) in Boyer and Hampton (2001) suggest that squid, epipelagic fish and lightfish and mesopelagic fish constitute a significant proportion of adult hake diet, but their main prey items are their own young and other demersal species. Shannon and Jarre-Teichmann (1999) assumed a diet of 50% macrozooplankton, 3% jellyfish, 1% cephalopods, 1% anchovy, 6% hake, 6% gobies, 4% macrobenthos, 18% mesopelagic fish, 10% other demersal fish and 1% horse mackerel. However, this was for adults and juveniles combined. Similar to the juvenile hake, the estimates given by Huse et al. (1998) for M. capensis (which were mostly larger than 50 cm) were used in conjunction with Pillar and Barange (1997) and Andronov (1983) to disaggregate the mixed contents and miscellaneous groups given in Huse et al. (1998). The adult diet was broken down into 26.9% euphausiids, 4.5% mesopelagics, 17.9% cannibalism, 6% demersals, 4.5% cephalopods, 4.5% horse mackerel and 4.5% each for macrobenthos, jellyfish, copepods, small pelagics, anchovy, sardine and snoek. To balance and fit the model the diet of adult hake had to be changed substantially. The 4.5% snoek in the diet of adult hake was reduced to 1%, 0.01% added to linefish, 1% to adult horse mackerel and the remaining 3.499% to macrozooplankton. Anchovy was reduced from 1.4% to 1% to fit and balance this group. Similarly, mesopelagics and juvenile horse mackerel were reduced from 4.5% to 1% each and 1% was added to adult horse mackerel. Cannibalism of juveniles by adults was reduced to 0.5% and added to adults cannibalizing themselves. The 6% demersals was reduced to 0.5% and monk (0.05%) and macrozooplankton (4.5%) were added. Cephalopods were reduced from 4.5% to 0.5% and macrozooplankton added. 20. Monkfish There are two species of monkfish in Namibia, namely Lophius vomerinus and L. vaillanti (Maartens and Booth, 2005). L. vomerinus is the more important species and occurs from northern Namibia (21°S) to the east coast of South Africa, while L. vaillanti occur north of Walvis Bay (Maartens and Booth, 2005). There are two separate spawning areas for L. vomerinus, one off Walvis Bay between 150 and 300 m depth, and one near the Orange River, at depths of 100–300 m (Maartens and Booth, 2005). Hampton (2003) showed that monkfish are caught mainly offshore, with large aggregations from the Angola-Benguela front extending south to 30°S, although there did seem to be interruptions at Sylvia Hill (25°S) and around the Orange River (29°S). Abelló et al. (1988) found an interruption in the monkfish stock at 27°S, around Elizabeth Bay, and a smaller reduction at Sylvia Hill. However, in this model the monk was assumed to occur on the whole coast, from the slope to the deep ocean. Catch statistics for monkfish date back to 1974 when the two species were taken as bycatch in the hake fishery (Hampton, 2003). However, the Sea Around Us database gives catches back to 1967, and the Sea Around Us catches from 1967 to 1977 and 2001 to 2003 were used in conjunction with the catches from Willemse (2002) for the time period 1978–2000 (Figure 18). Maartens and Booth (2001) give estimates of the effort by the monkfish and sole fishery in days fished from 1991 to 1999.  49  18  200  16  180  14  160 140  12  120  10  100  8  80  6  60  4  40  2  20  0 1967  Biomass ('000 t)  Booth and Quinn (2006) use an age-structured maximum likelihood model to estimate the exploitable biomass of monkfish (L. vomerinus) from 1975 to 2004 and that was used as a biomass trend (Figure 18) for this species. To fit and balance the model a biomass of 0.6 t·km-2 was needed. For a P/B ratio the natural mortality of both species were obtained from Fishbase (at 8°C), which was 0.27 for L. vaillanti and 0.21 for L. vomerinus, giving an average P/B of 0.24 year-1. A P/Q of 10% was assumed, similar to other demersals (Shannon and JarreTeichmann, 1999).  Catch ('000 t)  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  0 1972  1977  1982 Catch  1987  1992  1997  2002  Biomass  Figure 18. Catch and biomass (*1000 tonnes) of monkfish in the northern Benguela ecosystem.  Diet of monkfish was given by Walmsley et al. (2005) for the west coast of South Africa, and according to them, smaller individuals are found in shallower depths and there is a large overlap in distribution of the three size classes (<37 cm, 37–45 cm, and >45 cm); thus all size classes are exposed to fishing. Data on the size structure of monkfish indicate that 26.8% of landed monkfish were <37 cm in monkfish-directed trawls and 10.7% in hake directed trawls (Walmsley et al., 2005). According to Macpherson (1985) in Boyer and Hampton (2001), hake is the main prey of monkfish and monkfish would take any hake up to their own size. The diet of monk was then estimated as 36.7% hake, 29.4% demersals, 6.6% horse mackerel, 0.2% gobies, 1% cannibalism, 12.3% small pelagics, 3.9% anchovy, 0.4% mesopelagics, 8.4% cephalopods and 1.2% macrobenthos (Walmsley et al., 2005). The 0.2% gobies was increased to 10.2% and other demersals reduced to 5% to balance that group while adult horse mackerel was added (5%), small pelagics (19.7%) and mesopelagics were increased to 12.9%. The 36.7% hake was divided into 0.1% juveniles and 26% adult hake and small pelagics (10.6%). The 8.4% cephalopods was reduced to 3.4% and 5% added to macrobenthos. 21. Other demersals The other groundfish species caught off Namibia (in order of abundance) include large eye dentex (Dentex macrophthalmus), grenadiers (Trachyrinchus scabrus), jacopever (Helicolenus dactylopterus), Atlantic greeneye (Chlorophthalmus atlanticus), Cape elephant fish (Josef) (Callorhinchus capensis), longfin bonefish (Pterothrissus belloci), Cape gurnard (Kaapse korhaan) (Chelidonichthys capensis), black slimehead (Hoplostethus cadenati), banded rattail (Coelorinchus fasciatus), grenadiers (Nezumia spp), thinlip splitfin (Synagrops microlepis), kingklip (cusk-eel) (Genypterus capensis), west coast sole (Austroglossus microlepis), silver scabbardfish (bottersnoek) (Lepidopus caudatus), Gunter’s cuskeel (Selachophidium guentheri) and pencil cardinal (Epigonus denticulatus) (Hamukuaya et al., 2001). To this group is also added the orange roughy (Hoplostethus atlanticus) and alfonsino (Beryx splendens). The groundfish assemblages of the northern Benguela have been divided into shelf and slope assemblages that were then subdivided into northern and southern assemblages as well (Hamukuaya et al., 2001). However, for this project, as demersal fish are mostly combined in this group, they were assumed to occur everywhere in the ecosystem.  50  90  500 450 350  60  300 250 200  30  150  Biomass ('000 t)  400 Catch ('000 t)  The relative abundance of the demersal species for 1990–2000 was estimated by Bianchi et al. (2001) and the biomass for demersals from 1983 to 1990 was obtained from Macpherson and Gordoa (1992). Thus, using the average biomass for 1990 (144,000 tonnes) obtained from Macpherson and Gordoa (1992) and the catch rate from Bianchi et al. (2001), a time series of biomass from 1983 to 2000 was obtained (Figure 19). However, no biomass was available for 1956 for this group and biomass was thus estimated by assuming an ecotrophic efficiency of 95%. Catches (Figure 19) for all demersals were obtained from the Sea Around Us database.  Northern Benguela Ecosystem, Heymans and Sumaila  100 50 0 1950  1960  1970  1980  Catch  Estimated Biomass  1990  0 2000  Biomass  Figure 19. Catch and biomass estimated for demersal fish in the northern Benguela ecosystem (*1000 tonnes).  Shannon and Jarre-Teichmann (1999) assumed P/B and P/Q ratios of 1.0 year-1 and 10% respectively and a diet of 8.3% copepods, 32.4% euphausiids, 5.7% small pelagics, 1.9% myctophids, 2.2% gobies, 10.6% hake, 10.6% cannibalism, 13.3% cephalopods and 15% benthos. The 10.6% cannibalism was divided into 5% cannibalism, 1% monkfish, and detritus. The 10.6% hake was divided into 0.1% juvenile hake and 1% adult hake with the rest added to mesozooplankton. The 13.3% cephalopods was reduced to 0.01% and the rest added to detritus. The 15% macrobenthos was removed and 1% each added to lobster and crabs, with the rest added to detritus. 22. Cephalopods The important cephalopods include Loligo vulgaris reynaudii, Todarodes angolensis, Todaropsis eblanae, Lycoteuthis diadema, Sepia australis, Octopus spp. and Argonauta spp. (Lipiński, 1992). In addition the flying squid (Todarodes sagittatus) and cuttlefish (Sepia australis) were found amongst the demersal trawls (Hamukuaya et al., 2001). Abelló et al. (1988) showed that T. sagittatus occurred in deeper water between 23°S and 29°S, with the highest numbers around 24°S–25°S and 27°S. Similarly, Hamukuaya et al. (2001) showed that T. sagittatus occurred in demersal deep-water slope assemblages between 17°S and 29°S and that S. australis occurred on the shelf around the Orange River (29°S). Thus, as this group encompasses all cephalopods, it was assumed to occur everywhere in the spatial model. Loligo v. reynaudii and T. angolensis sustained fisheries in the 1980s, with catches obtained from Lipiński (1992) for 1980–1988 and biomass estimated from 1983 to 1988 for T. angolensis specifically. These estimates were for a smaller area and in winter, so the biomass was lower than the total catch; thus the estimates were not used in the model, but biomass was estimated by assuming an ectrophic efficiency of 95%. For the periods 1950–1979 and 1989–2003 catches were estimated as 1.8% of the main catch of all demersal species including hake, other demersals and monkfish as suggested by Hamukuaya et al. (2001). The P/B and P/Q ratios were 1.5 year-1 and 10% respectively (Shannon and Jarre-Teichmann, 1999). To fit the model, the catch of cephalopods in the first year had to be halved to 0.002 t·km-2·year-1. The diet of T. angolensis was calculated by Lipiński (1992) to include 5–10% euphausiids, 5–10% lightfish, 50–64% lanternfish, 5–8% hake, 2% other fish, 10% cephalopods, and 1% other organisms, while Shannon and Jarre-Teichmann (1999) included horse mackerel and gobies in the diet of all cephalopods. Therefore this model used a preliminary diet of 7.5% euphausiids, 64.5% lanternfish, 6.5% hake, 10% cannibalism, 5.5% horse mackerel and 6% gobies. The 6% gobies was increased to 10% to balance and the 64.5% of mesopelagics was reduced to 10% to balance that species. The difference was obtained from gobies, macro(20%) and mesozooplankton (24.5%). Juvenile horse mackerel was reduced to 1% and the remainder added to macrozooplankton. The 6.5% juvenile hake was reduced to 0.1% and the remainder added to macrozooplankton. The 10% cannibalism was reduced to 1% and added to gobies and mesozooplankton.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  51  23. Benthos The biomass of benthos was estimated by using an ecotrophic efficiency of 95%. The P/B (1.2 year-1) and P/Q (12%) of benthos in the southern Benguela was used (Jarre-Teichmann et al., 1998) and the diet composition of 5% benthic producers, 15% benthos and 80% detritus was obtained from Shannon and Jarre-Teichmann (1999). The 15% benthos was reduced to 5% macrobenthos, 0.1% crabs and 0.1% lobster and the remainder added to detritus. As this was a combined group for all benthos, benthos were assumed to occur in all areas of the spatial model. 24. Crab Deep-sea red crab (Chaceon maritae) occurs on the continental slope from 27°S northwards into Angola (Dias and Seita Machado, 1974 in Maurihungirire, 2004) at depths of 300–900 m (Melville-Smith, 1983). Thus, in the spatial model crabs were assumed to occur in the deep waters and on the northern slope. The fishery for deep-sea red crab started in 1973 off Namibia, catching 3,877 metric tonnes (Beyers and Wilke, 1980) and expanded rapidly, with the largest annual catch of 10,000 tonnes being recorded in 1983 (Le Roux, 2001). Fishing effort increased in 1974 with 17 Japanese vessels and one mother vessel exploiting the stock, but during the next years the effort decreased and by 1979 only 5 vessels were active (Beyers and Wilke, 1980). Catch estimates (Figure 20) from 1980 to 1997 were obtained from Boyer and Hampton (2001) and from 1998 to 2002 from Nicols (2004). The catch estimate for 1973 (Beyers and Wilke, 1980) was used in conjunction with the catch in 1980 to calculate a linear increase in the catch for 1974 (the Sea Around Us catch was smaller than that of 1973 and the increased effort would have indicated an increased catch). Catch estimates from 1975 to 1979 and 2003 given by the Sea Around Us project were used.  Catch ('000 t)  The crab biomass (Figure 20) has 45 12 declined from about 40,000 tonnes in the early 1980s to around 10,000 tonnes in the 1990s 9 (Hampton, 2003). Thus, an 30 estimate of 40,000 tonnes (0.22 t·km-2) was used as a first biomass 6 estimate for crabs in the 1950s. 15 Population models, supplemented 3 by surveys (which were conducted twice a year until 2000), indicate that the 2001 stock size was 0 0 around 10,000 tonnes whole mass 1970 1975 1980 1985 1990 1995 2000 2005 (Hampton, 2003). Estimates of crab biomass from 1990 to 1995 Catch Biomass were obtained from the Ministry of Fisheries and Marine Resources Figure 20. Catch and biomass of deep-sea crabs (*1000 tonnes). (Heymans, 1997). The P/B (2.5 year-1) and Q/B (8.6 year-1) ratios were obtained from Heymans (1997). The P/B ratio was lowered to 1.2 year-1 to fit the model.  Biomass ('000 t)  In 1976 a South West African company entered the fishery, with 2 vessels in 1977 (Beyers and Wilke, 1980). The number of vessels in the crab fishery from 1973 to 1986 and the CPUE and effort from 1980 to 1986 were obtained from Melville-Smith (1988). In 1993 a depth restriction of 500 m was introduced, and it was reduced to 400 m later that year to prevent fishing vessels from operating at the shallow end of the species’ range (Le Roux, 2001). Catches of deep-sea crab from 1980 to 2000 were obtained from Boyer and Hampton (2001).  Deep-sea crabs feed on skates (Macpherson, 1985 in Boyer and Hampton, 2001) and deep-sea (demersal) fish such as Cottuncoloides macrocephalus and Alepocephalus rostratus (Macpherson, 1983 in Boyer and Hampton, 2001). According to Beyers and Wilke (1980) the stomach contents of deep-sea crabs consisted of small shell fragments from goose barnacles, fish scales and the bones of small fish, while the remains of small crustaceans and unidentifiable digested food were the major constituents. Thus, this model assumed  52  Northern Benguela Ecosystem, Heymans and Sumaila  a diet of 33% macrobenthos, 33% demersal fish and 33% detritus. The demersals were reduced to 1% to balance the model and 1% was added to monkfish with the remaining 31% going to detritus. 25. Lobster Lobster (Jasus lalandii) is associated with upwelling areas and occurs from South Africa to approximately 25ºS (Hampton, 2003). They are most abundant off Lüderitz and the adjacent rocky shores (Hampton, 2003). Lobster was exploited from the early 1920s, and annual catches remained relatively high (9,000 tonnes) until the mid-1960s, and by the 1967 the fishery was near collapse (Pollock et al., 2000). The fishery uses baited traps and hoop nets. Catches (Figure 21) of lobster from 1950 to 1957 were obtained from Crawford et al. (1989), from 1955 to 1957 from Pollock and Shannon (1987), from 1958 to 1989 from Boyer and Hampton (2001), from 1990 to 1997 from Pollock et al. (2000), from 1998 to 2002 from Nicols (2004) and for 2003 from the Sea Around Us database. 14 12  Catch ('000 t)  The fishable biomass of lobster has declined from around 10,000 tonnes in the early 1970s to 2,000–3,000 tonnes in 2001 (Grobler, 2000 in Hampton, 2003). Hampton (2003) suggested that the fishable biomass was between 3,000 and 4,000 tonnes in 2000. Barkai and Branch (1988) estimated the annual Q/B and P/Q of rock lobster to be 2.46 and 0.11 respectively. However, to fit the model a P/B of 1.2 year-1, similar to that of macrobenthos was assumed and a P/Q of 0.3 was used to get a Q/B of 4 year-1.  10 8 6 4 2 0 1950  1960  1970  1980  1990  2000  Figure 21. Catches of lobster (*1000 tonnes) made in the northern Benguela ecosystem from 1950 to 2002.  Rock lobster feed on mussels (ribbed mussels such as Aulacomya ater), echinoderms, gastropods, bryozoans, polychaetes and seaweed (Boyer and Hampton, 2001). They are preyed on by octopus, dogsharks, hagfish, young seals and cannibalism (Boyer and Hampton, 2001). This model used a diet of 50% macrozooplankton and 50% seaweed. 26. Juvenile jellyfish Venter (1988a) found that there were abundant and widely distributed jellyfish off the coast of Namibia since the decrease in the pelagic fishing in 1972 and that the most abundant species were Chrysaora hysoscella and Aequorea aequorea. He showed that the main concentrations of jellyfish were between Cape Frio and south of Walvis Bay and mainly inshore. However, there were also some jellyfish further south and north and offshore around 21.5°S and at Walvis Bay (Venter, 1988a). Fearon et al. (1992) found that jellyfish occur throughout the area they surveyed from 17°S to 29°S, with the lowest concentrations around 17°S–18°S and the highest concentrations of Aequorea in the north (19°S–22°S) and the highest concentrations of Chrysaora in the south (23°S–27°S). Sparks et al. (2001) found that both species are more prevalent in the north and Chrysaora specifically was very abundant in the north. Pagès (1992) showed that various cnidarians occurred in different areas of the northern Benguela, both offshore and onshore and in all the areas studied between 17.5°S and 26°S. Lynam et al. (2006) showed that Chrysaora hysoscella seems to occur mostly on the shelf and is widely distributed with hotspots around Möwe Bay (19°S), Swakopmund/Walvis Bay (22°S–23°S) and the Orange River (28°S–30°S). By contrast, Aequorea forskalea seems to occur further offshore, with definite hotspots at Cape Frio (18.5°S), a large area from Henties Bay (22°S) to south of Sandwich Harbour (24°S) and at the Orange River mouth (28°S–30°S). Thus both adult and juvenile jellyfish were assumed to occur in all areas of the model except the deep ocean.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  53  Jellyfish was divided into adults and juveniles, and the P/B for juvenile jellyfish was estimated at 5 year-1. The diet was assumed to consist only of phytoplankton. 27. Adult jellyfish Venter (1988a) gave the distribution of jellyfish from the pelagic fish catches between 1981 and 1987. The catch of jellyfish from 1981 to 1987 was measured by Venter (1988a), who suggested that this was not an absolute estimate of catch, as it depends on the extent of damage caused during catching, transport and pumping procedures, on dumping and on the intensity of fishing activities. However, as the only estimate available, it was used as an indication of the fishing during that time. In addition, Venter (1988a) also indicates the percentage of the purse seine catch that consisted of jellyfish to be on average 14% for that time period. This model therefore assumes that from 1950 to 1972 the percentage of jellyfish in the catches increased linearly from 0% to 2% of the catches of anchovy and sardine (only), and from 1973 to 1979 it increased more substantially from 3% to 14% (the average for 1980–1988). From 1989 to 2003, we assumed that the jellyfish in the catch increased from 14% to 20% of the anchovy and sardine catches. These catches were still an order of magnitude higher than those given by Venter (1988a), so this was divided by 10 to get a first estimate of jellyfish catch (Figure 22). 5  60  3 40 2 20  1 0  Biomass (million t)  80  4  Catch ('000 t)  Pagès (1992) found that the most abundant medusae in the northern Benguela in 1981–1982 were Chrysaora hysoscella, Aequorea aequorea and Liriope tetraphylla, with the most abundant siphonophores being Muggiaea atlantica, Abylopsis tetragona and Bassia bassensis. He differentiated three assemblages, one associated with coastal upwelling, the second with oceanic water and the third with species typical of shelf fauna (Pagès, 1992). C. hysoscella is known as a coastal species and A. aequorea a coastal shelf species, while L. tetraphylla is oceanic, M. atlantica is an inshore species and Abylopsis tetragona is an oceanic species (Pagès, 1992).  0 1980  1985  1990 Catch  1995  2000  Biomass  Figure 22. Catch (*1000 tonnes) and biomass (million tonnes) estimates of jellyfish.  Estimates of jellyfish biomass (Figure 22) from 1982 to 1989 were obtained from Fearon et al. (1992), while the biomass for 1997 to 1998 was estimated by Sparks et al. (2001), and Lynam et al. (2006) estimated a biomass of 106 tonnes·km-2 in 2003. However, the biomass in the 1950s was not known, and according to Lynam et al. (2006) large jellyfish was not prevalent prior to the heavy exploitation of pelagic fish in the Benguela, as “reports of extensive plankton sampling in the 1950s and 1960s do not mention large jellyfish, although numerous small gelatinous species (e.g. ctenophores) were observed”. We can therefore assume that the biomass would have been very low in the 1950s and a value of 13.2 tonnes·km-2 was needed to fit the model. The P/B and P/Q ratios for adults were estimated at 0.371 year-1 and 41% respectively (Shannon and JarreTeichmann, 1999). The diet of adults was estimated to consist of 24.7% phytoplankton, 25% bacteria (detritus), 25% microzooplankton (detritus) and 25% copepods (Shannon and Jarre-Teichmann, 1999) and we assumed that juvenile anchovy, sardine and gobies were 0.1% each. 28. Macrozooplankton Early studies of the zooplankton of the northern Benguela showed that the area west and northwest of Walvis Bay had consistently high zooplankton biomasses and that there are two annual zooplankton  54  Northern Benguela Ecosystem, Heymans and Sumaila  peaks, one during the late spring and early summer (November–December) and the other during the autumn (March–May) (Shannon and Pillar, 1986). The SWAPELS cruises found that there were high concentrations of phytoplankton inshore and offshore peaks of zooplankton biomass (Shannon and Pillar, 1986). Estimates of zooplankton biomass were given in Shannon and Pillar (1986) for 1958–1959 as 2.04 gC·m-2 and for 1959–1962 as 1.34 gC·m-2 in Walvis Bay, for the area 19ºS–23ºS in 1971 as 0.96 gC·m-2 and in June–July as 0.63 gC·m-2. This averages to about 18 t·km-2 for both meso- and macrozooplankton combined (Shannon and Pillar, 1986). However, the macrozooplankton biomass was estimated using an ecotrophic efficiency of 95% to balance the model. Macrozooplankton include euphausiids, chaetognaths and hyperiid amphipods, with the euphausiids being the most important group (Heymans, 1997). Amphipods, siphonophores, ostracods, decapods and mysids are scarce (Timonin et al., 1992). The important euphausiids include Nyctiphanes capensis, Nematoscelis megalops, Thysanoessa gregaria, Euphausia lucens, E. tenera, E. recurva and E. hanseni (Hart and Currie, 1960). Species of chaetognaths of importance include Sagitta friderici, S. serratodentata, S. decipiens, S. minima, S. lyra and Eukrohnia hamata (Hart and Currie, 1960). In addition, Hart and Currie (1960) also found Cladocera (Evadne normandi, Podon polyphemoides), Ostracoda (Conchoecia spp.), mysids (Gastrosaccus sanctus, Boreomysis rostrata), Cumacea (Iphinoe spp., Upselaspis caparti), Stomatopods (Squilla armata), planktonic mollusks (Ianthina ianthina, I. globosa) and Larvacea (Oikopleura and Fritillaria). Hutchings et al. (1991) estimated a biomass of 0.6 gC·m-2, or 15 t·km-2 (using a C:wet weight ratio of 0.04) and P/B and P/Q ratios 13 year-1 and 41% respectively, while their diet included 60% phytoplankton and 40% copepods. 29. Mesozooplankton The mesozooplankton of the northern Benguela is homogeneous, with copepods being dominant in terms of numbers and biomass (Timonin et al., 1992). Hansen et al. (2005) found that 4 calanoid copepods dominated the zooplankton community off Walvis Bay in 2000, ranked in order of abundance as Metridia lucens, Calanoides carinatus, Rhincalanus nasutus and Centropages spp. The total copepod abundance was 520*103 m-2 and according to Hansen et al. (2005) it compared well with the 236–446*103 m-2 found in 1972, 700*103 m-2 in 1979 and 300–800*103 m-2 in 1986. Unterüberbacher (1964) in Shannon and Pillar (1986) gave an estimate of 0.828 gC·m-2 and Kollmer (1963) in Shannon and Pillar (1986) estimated 1.26 gC·m-2; thus using the average of these two estimates and the carbon to wet mass ratio of 0.04 given by Hutchings et al. (1991) gives a biomass of 26 tonnes·km-2 between 1958 and 1962. This figure is used here and compares to the 25 tonnes·km-2 used by Shannon and Jarre-Teichmann (1999). This is also very similar to the value of 1.0 gC·m-2 estimated by Hutchings et al. (1991). The P/B and P/Q ratios were 40 year-1 and 30% respectively (Hutchings et al., 1991). However, Cushing (1971) in Verheye et al. (1992) suggests a daily P/B of 0.02–0.03 or 9.1 year-1. The diet of copepods includes 50:50 phytoplankton and detritus in our model (Hutchings et al., 1991). 30. Benthic producers Benthic producers in Namibia include the seaweed Gracilaria gracilis, which is harvested in Lüderitz Bay and the lagoon area and is used for the production of agar (Hampton, 2003). Kelp species such as Laminaria pallida and L. schinzii and Ecklonia maxima are also prevalent in the northern Benguela and used to a lesser extent (Hampton, 2003). Critchley et al. (1991) found that Suhria vittata is also exploited for export. The organized harvesting of seaweed started in 1980 and beach-cast Gracilaria was exported for US$1,700 per tonne, but that declined to US$1,250 per tonne by 1994. Production increased from 310 tonnes dry weight in 1982 to 1,500 tonnes in 1988 (Hampton, 2003). Approximately 15,000 tonnes of Gracilaria are deposited on the shore in Lüderitz and collected for agar extraction each year (Critchley et al., 1991). An agar extraction plant was relocated to Lüderitz in 1986 and it employed about 250 people. By 1991 it had a targeted output of 100 tonnes of agar powder per year, which was exported to Europe, South Africa and Mauritius (Critchley et al., 1991). The agar powder has a moisture content of 10–16% and the agar content of Gracilaria verrucosa ranged from 16–30% of dry weight, with monthly yields of 5.3 tonne per month in 1988–1989, 7 tonne per month in 1989–1990 and 12 tonne permonth in 1990–1991 (Critchley et al., 1991). The total dry weight extracted from 1986 to 1989 was given by Critchley et al. (1991) as approximately 810 tonnes, 982 tonnes, 1,640 tonnes and 1,244 tonnes from 1986 to 1989. This model assumed a linear increase in the catch from 1982 to 1986.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  55  Harvests since 1990 have fluctuated between 423 and 1,029 tonnes dry weight, with the largest harvests being made in 1995 and 1996. We assumed that the 423 tonnes was in 1990 and the 1,029 tonnes in 1995 and 1996 and did a linear increase from 1990 to 1995. In 2000, 829 tonnes were harvested with an export value of N$3,850,000 (Hampton, 2003), and Nicols (2004) gives estimates of harvest from 1998 to 2002. A dry weight to wet weight ratio of 0.137 was used to calculate the wet weight (Hernández et al., 2005). We assumed that there was a linear increase in catch from 1979 to 1982 and from 1982 to 1988 and that there was a linear decline from 1996 to 1998. No estimates of biomass or P/B were available for the 1950s, thus the values from the 1970s were used for P/B (0.5 year-1) and an EE of 0.5. 31. Phytoplankton The dominant phytoplankton or ‘microplankton’ obtained during the William Scoresby surveys include: Stephanopyxis turris, Thalassiosira excentrica, T. subtilis, Planktoniella sol, Chaetoceros affine, C. compressum, C. constrictum, C. convulutum, C. debile, C. decipiens, C. difficile, C. didymum, Ceratium strictum, Rhizosolenia alata, R. hebetata, R. imbricata, Dactyliosolen mediterraneus, F. karstenii, Asterionella japonica, Nitzschia delicatissima, N. seriata, Peridium spp., Trichodesmium thiebautii and Foramanifera (Hart and Currie, 1960). During 1950 the William Scoresby undertook a major oceanographic study off the coast of southern Africa and found that the depth of upwelling was down to 350 m in the fall (March) and 320 m in spring (October) (Hart and Currie, 1960). Even during this survey there were large areas of oxygen-depleted water in the Walvis Bay region (Hart and Currie, 1960) and Brongersma-Sanders (1948) in Hart and Currie (1960) reported large areas of fish kills from Walvis Bay to Conception Bay, while Reuning (1925) in Hart and Currie (1960) showed even larger fish kills in 1924– 1925. These fish kills are related to calm weather or northerly winds, which create abnormal conditions in the vicinity of Walvis Bay (Hart and Currie, 1960). Barange and Boyd (1992) showed that the main areas of high chlorophyll-a distribution between 1985 and 1989 were south of Cape Frio (18.5°S) to about Torra Bay (20.5°S) and from Swakopmund (22.5°S) to about Easter Point (25.5°S). These overlap with the areas of highest chlorophyll-a concentration shown by Pitcher et al. (1992) for 1971–1989, with highest concentrations between Cape Cross and just south of Cape Frio (21.5°S–19°S) from Swakopmund to Sandwich Harbour (22.5°S–23.5°S) and from 24.5°S north of Sylvia Hill to Lüderitz (26.5°S). These areas are mostly north of the main upwelling areas around Lüderitz (from 24°S–26.5°), north of Cape Cross (20°S–21.5°S) and around Cape Frio (18°S–19°S), according to Crawford et al. (1989). The upwelling cells were also given by Hamukuaya et al. (2001) from around Panther Head (28°S) to Dolphin Head at Easter Point (25.5°S), north of Cape Cross (20°S–21.5°S) and south of Cape Frio (18.5°S–19°S), indicating that the northern upwelling cell might have moved. In addition, Pitcher et al. (1992) showed that the highest offshore concentrations of chlorophyll-a were from the Kunene (17°S, and their most northerly data) to Cape Cross (21.5°S) and between 24°S and 28°S. Wasmund et al. (2005) calculated the primary production and chlorophyll-a biomass of phytoplankton in 6 different water bodies in the Cape Frio area, and the average primary production was 1.6 gC·m-2·d-1 (24,436 tww·km-2·year-1) and 56 mg·m-2 (117 tww·km-2) respectively. Carr and Kearns (2003) give estimates of the primary production and biomass of phytoplankton between 1998 and 2000 for 6 areas off Namibia, with an average primary production of 4.25 gC·m-2·d-1 (65,153 tww·m-2·year-1) and chlorophyll biomass of 6.83 mgC·m-3. For 1997–1999 Carr (2002) estimates a production of 2.49 gC·m-2·d-1 (40,057 tww·km-2·year-1) and Ryther (1969) estimates a value of 3.9 gC·m-2·d-1 (59,787 tww·km-2·year-1) prior to 1969. Estimates of primary production given by references in Shannon and Pillar (1986) range from 0.8 gC·m-2·d-1 at the Orange River to 2.5 gC·m-2·d-1 at Walvis Bay to 5.1 gC·m-2·d-1 at Walvis Bay, to 0.3 gC·m-2·d-1 at Orange River, 0.9 gC·m-2·d-1 at Lüderitz, 3.9 gC·m-2·d-1 at Sylvia Hill, 1.2 gC·m-2·d-1 at Walvis Bay, 0.6 at 21ºS and 3.2 gC·m-2·d-1 at Cape Frio. This gives an average similar to the 2.58 gC·m-2·d-1 of Schultz (1982) in Wasmund et al. (2005). In addition, Schultz (1982) in Wasmund et al. (2005) estimated a chlorophyll-a biomass of 258 mg·m-2, or 433 t·km-2 using a C:wet weight ratio of 42 (Cushing et al., 1958). The biomass was estimated using an average euphotic zone depth of 40 m, as Wasmund et al. (2005) suggests that the 1% light level was at 18–36 m in upwelling water and 30–60 m in subtropical waters, and the average depth of euphotic zones of the various water bodies they tested was approximately 40 m. This value is lower than the 600 t·km-2 estimated by Brown et al. (1991). The annual P/B ratio of 49.9 was estimated from the biomass and primary production (Schultz, 1982 in Wasmund et al., 2005). This is much higher than the annual P/B of 30 estimated by Brown et al. (1991).  56  Northern Benguela Ecosystem, Heymans and Sumaila  Environmental variation Shelton and Crawford (1988) found that three major components of variability with different time periods characterize the Benguela system, namely upwelling events, seasonal changes and prolonged changes in normal advection patterns. The relative contribution of these different components to the overall variability of the system has not been resolved (Shelton and Crawford, 1988). Benguela Niños occurred in 1934, 1949, 1963, 1974, 1984 (Shelton and Crawford, 1988) and 1995. These warm events are interspersed with cooler periods, suggesting a 10-year cycle, and correspond with heavy rains or flooding in parts of Namibia (Shannon et al., 1986). The rainfall patterns in Namibia have an effect on the oxygen content of the bottom water along the Namibian coast. It has been postulated that groundwater with high silicate concentrations seeps into the coastal water at river mouths (Pollock et al., 2000). In adjacent river catchments the rainfall regime changed from the 1960s. It increased during the late 1960s and the 1970s but decreased again during the 1980s and 1990s (Pollock et al., 2000). Thus, a nutrient-driven change in ecosystem productivity might have started in the late 1960s with the consequence of increased production of oxygen-depleted bottom water in the area between 19°S and 25°S. This was synchronized with the collapse of the west coast pilchard stock and its replacement by anchovy and pelagic gobies (Crawford et al., 1985; Pollock et al., 2000). The recycling of silicate over the siliceous mudbelt between 19°S and 25°S might be maintaining the system in a hypertrophic mode with persistently high levels of diatom production and decay (Pollock et al., 2000). Rainfall estimates from 1913 to 1997 were obtained from Pollock et al. (2000). Two environmental anomalies occurred between 1993 and 1995: in 1993–1994 the shelf water off Namibia was hypoxic and extended over a larger area and lasted for a longer period than usual (Hamukuaya et al., 1998). Then in February–March 1995 there was a strong Benguela Niño event, with warm water intruding onto the Namibian shelf (Cury and Shannon, 2004). Most stocks underwent large declines: sardine recruitment was poor (Boyer et al., 2001a), hake migrated offshore (Hamukuaya et al., 1998), monkfish catches decreased (Maartens, 1999 in Cury and Shannon, 2004), horse mackerel biomass decreased by 50% (Boyer and Hampton, 2001) and anchovy abundance declined further (Cury and Shannon, 2004). Warm water intrusions into northern Namibia also have an effect on the ecosystem. Voges et al. (2002) give indications of how far south warm water intruded into Namibia from Angola, and Stander and De Decker (1969) in Shelton and Crawford (1988) indicated that in 1963 there was an intrusion of warm water from the northwest. There is evidence that the distribution of hake in Namibian waters is affected by both oxygen deficiency and temperature, although adult hake are well adapted to low oxygen concentrations (Hampton, 2003). Oxygen levels are generally low over much of the Namibian shelf due to its depletion during the decomposition of excessive phytoplankton production and the advection into the region of oxygendeficient water, with concentrations below 2 ml·ℓ-1, formed off Angola (Shannon, 1985). Oxygen levels over a large part of the shelf can become intolerable, possibly causing major shifts in distribution and even increased mortality, if extensive and persistent enough. The juveniles, which are less tolerant to hypoxic conditions, are particularly vulnerable (Woodhead, et al. 1996 in Hampton, 2003). An example of a major effect on the resource occurred in 1993 and 1994, when pronounced and persistent hypoxic conditions off central and northern Namibia displaced M. capensis offshore, subjecting them to heavy mortality from predation by larger hake and trawling (Hamukuaya et al., 1998). There is evidence to suggest that low surface temperatures favor hake recruitment, or at least that the recruits are more abundant and occur at higher densities in cool periods, such as in 1992, when M. capensis recruitment in Namibia was the highest ever recorded (Stromme, 1996 in Hampton, 2003). Also, there has been a clear positive correlation between monthly catch rates and surface temperature in Namibia in certain years (e.g. 1994 through 1996), although in other years (e.g. 1993 and 1997) the correlation has been as clearly negative (Boyer et al., 1998 and Gordoa et al., 2000 in Hampton, 2003). The reasons for these apparent correlations are not understood. The vertical structure in temperature and oxygen concentration, particularly close to the bottom, is believed to have an effect on the vertical distribution of hake, which could indirectly affect their horizontal distribution (S. Sundby, pers. comm.) and perhaps even their abundance. The effect on trawl catch rates, and hence estimates of abundance, would be more direct, since even a small vertical migration away from the bottom will affect catch rates significantly (Hampton, 2003).  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  57  Anchovy catches were exceptionally good in 1987 due to the breakdown of the Lüderitz upwelling barrier, which enabled anchovy recruits from South Africa to be carried into the northern Benguela (Hampton, 2003). According to Jenkins (1979) in Shelton and Crawford (1988), lower SST, lower sea level and lower SE wind components were positively correlated with pilchard catches. The rock lobster decline in the late 1980s was due to changes related to oxygen fluctuations in bottom waters and was aggravated by overfishing (Hampton, 2003).  0.50  1.00 0.00 0.96 -0.50  0.92 0.88 1956  -1.00 1966  Forcing function 1.08 1.04  1976 SST  1986  1996  10 Yr moving average (SST) 30  15  1.00 0 0.96 0.92  Sea surface temperature  Forcing function  1.04  Wind stress  MODEL FITTING RESULTS  1.00  1.08  Forcing function  Time series of wind stress for the Lüderitz upwelling cell from 1960 to 2002 were obtained from Klingelhoeffer (2006) and from 1906 to 1984 for the area 20ºS–25ºS:5ºE–10ºE from Shannon and Taunton-Clark (1988). Similarly for the same time period, sea surface temperature for the same area and for the whole coastal area were also obtained from Shannon and Taunton-Clark (1988). Sea surface temperature and pseudo–wind stress anomalies were obtained from Shannon and O’Toole (2003) from 1946 to 1990 and the sea surface anomaly was negatively correlated with the forcing function estimated by the model (Figure 23).  -15  To fit the model, the 0.88 -30 catches of tuna, seaweed 1956 1966 1976 1986 1996 and lobster were forced thereby acting as a simple stock reduction model for Forcing function Wind stress 10 Yr moving average (Wind) those groups. For sharks, anchovy, monkfish, Figure 23. Estimates of the forcing function estimated by the model based on demersals and crabs only the data : Top graph - Sea surface temperature (SST) anomaly. Bottom graph – some of the catches were wind stress. forced where no biomass was available to estimate a fishing effort. For the species that were not fished in 1956 the catches in their first year were entered and then taken out at a negative biomass accumulation to start the effort calculations correctly. In addition, to fit the model the catch of cephalopods in the first year had to be halved to 0.002 t·km-2·year-1 and a catch and discard of 0.002 t·km-2·year-1 needed to be added to juvenile horse mackerel to increase their catches in the fitted model. The input variables of the balanced model are given in Table 4, the catches used for the 1956 Ecopath model in Table 5 and the diet in Appendix A. The model was driven by the effort time series given in the fisheries section above, and the model was fitted to the time series by the algorithm that changes the vulnerabilities for each fo the interactions between predators and prey and by using feeding time adjustment rates of 0.5 for juvenile anchovy, sardine, gobies, and juvenile and adult jellyfish (Table 4). In addition a forcing function was estimated  58  which was negatively correlated (R2=-0.434, α=0.05, DF = 33) with the sea surface anomaly given by Shannon and O’Toole (2003) and the summer sea surface temperatures at Lüderitz (R2=-0.426, α=0.05, DF = 39) used in Heymans (2004). The forcing function was also positively correlated to the wind stress anomaly (R2=0.341, α=0.05, DF = 43) used by Klingelhoeffer (2006). The sea surface temperature anomaly and wind stress as well as the forcing function estimated by the model are given in Figure 23. Even though both these environmental variables were correlated with the forcing function obtained from the model, neither were  Northern Benguela Ecosystem, Heymans and Sumaila  Figure 24. Fit of the northern Benguela model (lines) to the biomass estimates given in the literature (dots).  directly applied to the model as they do not span the whole time period. The fitted model results are given in Figure 24 (biomass) and Figure 25 (catch). The model was able to reproduce the general decline in pelagics, specifically anchovy, quite well, but was not able to reproduce the large increase in sardine in 1960 that was estimated by VPA (Figure 24). This could indicate more of a failure of the VPA model than of the ecosystem model, as VPA can overestimate the initial biomasses in order to have large catches (Figure 25). The good fit of the hake and monkfish biomass in the model to the data are related to the high catches in the system, and the model  Figure 25. Fit of the model (lines) to the catches of the various species in the northern Benguela ecosystem.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  59  was able to reproduce the catches of those two groups very well, as well as the catches of other demersal fish species and crabs (Figure 25). Where the model is unable to fit the data is in the gobies, small pelagics and mesopelagics, where we have very little information on the biomass of these groups to drive the model. For the predators such as seals, sharks and snoek the model reproduces the catch series quite well even when we do not have good biomass estimates. Table 4. Ecosystem groupings, trophic level (TL), biomass (t·km-2) production to biomass (P/B) and consumption to biomass (Q/B) ratios, ecotrophic efficiencies (EE), catches (t·km-2·yr-1) and vulnerabilities (Vul.) by predator used in the northern Benguela ecosystem in 1956. Values in italics have been estimated by Ecopath. Group name TL Biomass P/B Q/B EE Catch Vul. Marine Mammals 4.23 0.027 1 12.66 0.030 2.00 Seals 3.84 0.150 0.08 19.44 0.890 0.005 2.00 0.838 2.00 Birds 3.54 0.017 0.16 120.3 2 0.577 0.0002 2.00 Sharks 4.09 0.411 0.5 5 0.005 0.005 1.50 Tuna 3.9 1.000 0.25 0.664 0.25 2.5 0.950 0.017 5.60 Snoek 3.66 Other linefish 3.74 0.121 0.2 2 0.585 0.007 1.00 0.704 1000.00 Anchovy juvenile 2 0.005 10 103.9 0.785 0.003 2.05 Anchovy adults 2.76 5.700 0.85 11.6 0.378 10.00 Sardine larvae 2 0.023 10 48.45 0.398 1.722 1000.00 Sardine adults 2.43 26.000 0.53 5 0.631 2.00 Gobies larvae 2 0.004 7 144.2 0.877 0.007 5.13 Gobies adults 2.09 6.300 1.1 15 9.206 0.75 7.5 0.950 0.079 2.00 Other small pelagics 3.17 3.820 1.23 12.3 0.950 0.007 1.00 Mesopelagics 3.24 0.821 0.152 2.00 Juvenile horse mackerel 3 0.934 1.7 15.18 0.390 0.007 2.00 Adult horse mackerel 3.34 7.000 0.4 5.2 0.891 2.00 Juvenile hake 3.38 0.482 1.05 11.27 0.407 0.134 2.00 Adult hake 3.35 24.000 0.4 2.4 2.4 0.787 0.001 1.75 Monkfish 4.02 0.600 0.24 1.158 1 10 0.950 0.073 2.00 Other demersals 2.78 1.023 1.5 15 0.950 0.002 2.00 Cephalopods 3.43 7.826 1.2 10 0.950 2.00 Macrobenthos 2.06 0.210 1.2 8.6 0.950 0.022 1.50 Crabs 2.4 0.235 1.2 4 0.950 0.043 2.00 Lobster 2.53 Jellyfish juveniles 2 0.002 5 11.16 0.354 2.00 0.826 0.001 3.00 Jellyfish adults 2.25 13.200 0.37 0.90 13.302 13 32.5 0.950 2.00 Macrozooplankton 2.4 133 0.316 2.00 Mesozooplankton 2 26.000 40 0.584 15 0.500 0.006 Benthic producers 1 0.101 Phytoplankton 1 433.44 49.9 0.087 Detritus 1 1.000 -  Ecospace model Base map and habitat types An Ecospace model of the northern Benguela, off the coast of Namibia (15°S–29°S, 10°E–17°E) was constructed to look at the effectiveness of current MPAs and the possible placement and number of MPAs in the future. The base map was obtained from the global GIS database query in Ecospace with a resolution of 6 minutes. The base map had 140 rows and 70 columns with a cell length of 10 km and 2 steps per degree (Figure 26). The base map depths were then exported to Excel and divided into 11 specific habitats based on depths <30 m as surf zones, 30–200 m as shelf areas, 200–500 m as slope areas and >500 m as deep areas. In addition the northern and southern parts of the coast were split at the Walvis Ridge. There are 3 marine protected areas in the northern Benguela, namely the Skeleton Coast Park in the north, the Namib Naukluft Park south of Walvis Bay, and the Diamond Sperrgebiet, south of Lüderitz (Figure 26). These areas are only closed to fishing from the shore and in the surf zone, but fishing in areas deeper than 30 m is allowed at present. The ports of Walvis Bay and Lüderitz are both used by the purse seine and demersal trawlers, while the lobster fishery and seaweed harvesting only occur around Lüderitz and the seal fishery only at Cape Cross. The midwater trawlers, longliners and crab fisheries are all operated from distant ports and with freezer trawlers. Linefish are caught by the commercial linefish fleet  60  Northern Benguela Ecosystem, Heymans and Sumaila  mostly from Walvis Bay, Swakopmund and Henties Bay (just north of Walvis Bay) and by the recreational fishery around Lüderitz and in the West Coast Recreational Area between Walvis Bay to Cape Cross. Table 5. Catches (t·km-2·yr-1) by the purse seine (PS), midwater (MWT), demersal (DT), longline (LL), crab, lobster, commercial line (Com L), recreational line (Rec L), seal and seaweed (SW) fisheries. Italics are discards vs. landings. Group name PS MWT DT LL Crab Lobster Com L Rec L Seal SW Seals 0.0005 0.0003 0.004 Sharks 0.0001 0.00002 0.00002 0.0001 0.000001 Tuna 0.004 0.0003 0.0004 Snoek 0.00001 0.017 Other linefish 0.002 0.0046 Anchovy adults 0.003 Sardine adults 1.722 Gobies adults 0.007 Other S pelagics 0.079 Mesopelagics 0.007 0.0002 J. h mackerel 0.152 A. h mackerel 0.0045 0.002 Juvenile hake 0.1340 Adult hake 0.0005 Monkfish 0.0727 Other demersals 0.0020 Cephalopods Crabs 0.022 Lobster 0.0434 Jellyfish adults 0.001 Benthic prod. 0.006 Total 1.976 0.007 0.210 0.0003 0.022 0.043 0.019 0.005 0.004 0.006  Habitat information was available for some species and fisheries from the literature and these are described in the compartment descriptions in the text above. The habitat preferences of the different species are given in Table 6 and the various fisheries have been defined by area in Table 7. This information was used to define the habitats used in this model, which included the: 1) Deep areas north of the Walvis Ridge (>500 m, 15.2% of the area); 2) Northern slope (200–500 m, 1.6% of the area); 3) Shelf (30–200 m, 13.4%); 4) Northern surf zone (<30 m, 0.3%); 5) Cape Cross (<0.1%); 6) West Coast Recreational Area (0.3%); 7) Southern surf zone (< 30 m, 0.3%); 8) Southern slope (200–500 m, 10.9%); 9) Deep areas south of the Walvis Ridge (>500 m, 57.9%); 10) Lüderitz harbour (<0.1%).  1  4 Skeleton Coast Park 2 6 WCRA  Cape Cross 5  3  Walvis Bay 9  Lüderitz 10  8  Namib Naukluft Park Sperrgebiet 7  Dispersal Figure 26. Base map of the northern  The dispersal parameters in Ecospace were used to confine Benguela ecosystem with numbers species to their known habitat. For most species the base defined in the text. dispersal rate was kept at the default value of 300 km·yr-1, while the relative dispersal in bad habitat was left at the default value of 5 km·yr-1, 100 km·yr-1 for species that are more pelagic and have specific areas of occurrence (adult and juvenile anchovy, sardine, gobies and juvenile horse mackerel) and 1,000 km·yr-1 for crabs and lobsters because they only occur in very specific areas in the ecosystem. In addition the relative vulnerability to predators in bad habitat was increased from the base value of 2 to 100 for juvenile and adult anchovy, sardine, gobies, as well as for juvenile horse mackerel, crabs and lobsters. Finally the relative feeding rate in bad habitat was reduced to 0 for these species from 0.01, which was the default for all other groups.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  61  Table 6. Preferred habitats for the various groups in the northern Benguela. CC = Cape Cross and WCRA = West Coast Recreational Area. Deep Slope Slope Deep Group All Shelf Surf N CC WCRA Surf S Lüderitz N N S S Marine mammals + Seals + + + Birds + Sharks + Tuna + + + + + Snoek + + + + + + + + Other linefish + + + + + + + + Anchovy juveniles + + + + + + Anchovy adults + + + + + + Sardine juveniles + + + + + + Sardine adults + + + + + + Gobies juveniles + + + + + + + + + + + + + + Gobies adults Other s pelagics + Mesopelagics + + + + + Juv. h mackerel + + + + + + Adult h mackerel + + + J hake + + + A hake + + + + Monkfish + + + + Other demersals + Cephalopods + Macrobenthos + Crabs + + Lobster + + + Jellyfish juvs + Jellyfish adults + Macrozooplankton + Mesozooplankton + Benthic producers + + + Phytoplankton + Detritus + Table 7. Distribution of the fishery in the northern Benguela Ecospace model. CC = Cape Cross and WCRA = West Coast Recreational Area. Deep Slope Slope Deep Fleet All Shelf Surf N CC WCRA Surf S Lüderitz N N S S Purse seine + Midwater trawler + + + + + Demersal + + + + + Longlines + + + + + Crab traps + Lobster + Commercial line + + + + + Recreational line + + + + + Seal fishery + Seaweed +  Advection Advection patterns are given by Carr and Kearns (2003) and currents by Stenevik et al. (2003). Phytoplankton and mesoplankton were advected in the model. In general, the wind vectors on this part of the coast are from north to south, with northwesterly winds being prevalent offshore. The wind forcing varies between 0.08 and 1.2 m2·s-1 (Carr and Kearns, 2003). Due to the Coriolis force, sea surface currents move the water offshore and upwelling occurs. The spatial model of the northern Benguela ecosystem is given in Figure 27.  62  Northern Benguela Ecosystem, Heymans and Sumaila  Figure 27. Result of the Ecospace run for biomass in the northern Benguela ecosystem.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  63  ACKNOWLEDGEMENTS The authors wish to thank Kevin Stephanus and Susan Alexander for their contribution to the data used in this report. The work was funded by the European Union through the INCOFISH project (www.incofish.org).  REFERENCES Abelló, P., Gordoa, A., Manriquez, M., Masó, M. and Macpherson, E. 1988. Biomass indices and recruitment levels for hake and other commercial species in ICSEAF Division 1.4 and 1.5: Results from the Spanish Benguela XI research cruise (1987). Collection of Scientific Papers of the international Commision of South East Atlantic Fisheries 15(1): 7-21. Andronov, V.N. 1983. Feeding of Cape horse mackerel (Trachurus trachurus capensis) and Cape hake (Merluccius capensis) off Namibia in January 1982. Collection of Scientific Papers of the international Commision of South East Atlantic Fisheries 10(1): 1-6. Andronov, V.N. 1985. Feeding of Cape horse mackerel Trachurus trachurus capensis Castelnau, in the Namibian area. Collection of Scientific Papers of the international Commision of South East Atlantic Fisheries 12(1): 1-16. Anon. 2001. Proceedings of an international workshop on the research and management of horse mackerel in Namibian waters. pp. 45++ in Payne, A.I.L., editor. Ministry of Fisheries and Marine Resources, Swakopmund, Namibia. Anon. 2005. A review of aquaculture policy and institutional capacity in the BCLME region, with recommended regional policy options. BCLME Project LMR/MC/03/01, EFA Enviro-Fish Africa, The Consortium 67 p. Armstrong, C.W., Sumaila, U.R., Erastus, A. and Msiska, O. 2004. Benefits and costs of the Namibianisation policy. pp. 203-214 in Sumaila, U.R., Steinshamn, S.I., Skog, M. and Boyer, D.C., editors. Ecological, economic and social aspects of Namibian fisheries. Eburon, Netherlands. Axelsen, B.E., Krakstad, J.-O. and Bauleth-D’Almeida, G. 2004. Aggregation dynamics and behaviour of the Cape horse mackerel (Tracurus trachurus capensis) in the northern Benguela - implications for acoustic abundance estimation. pp. 135-164 in Sumaila, U.R., Steinshamn, S.I., Skog, M. and Boyer, D.C., editors. Ecological, economic and social aspects of Namibian fisheries. Eburon, Netherlands. Balmelli, M. and Wickens, P.A. 1994. Estimates of daily ration for the South African (Cape) fur seal. South African Journal of Marine Science 14: 151-157. Barange, M. and Boyd, A.J. 1992. Life history, circulation and maintenance of Nyctiphanes capensis (Euphausiacea) in the northern Benguela upwelling system. South African Journal of Marine Science 12: 95-106. Barkai, A. and Branch, G.M. 1988. Energy requirements for a dense population of rock lobsters Jasus lalandii: Novel importance of unorthodox food sources. Marine Ecology Progress Series 50: 83-96. Barnes, J., Zeybrandt, F., Kirchner, C., Sakko, A. and MacGregor, J. 2004. Economic valuation of the recreational shore fishery: a comparison of techniques. pp. 215-230 in Sumaila, U.R., Steinshamn, S.I., Skog, M. and Boyer, D.C., editors. Ecological, economic and social aspects of Namibian fisheries. Eburon, Netherlands. Bauleth-D’Almeida, G. 2001. Development and application of management procedures for horse mackerel, Trachurus capensis, in the Namibian waters: 1990-2001. pp. 1-9 in Payne, A.I.L., editor. Proceedings of an international workshop on the research and managment of horse mackerel in Namibian waters. Ministry of Fisheries and Marine Resources, Swakopmund. Bauleth-D’Almeida, G., Krakstad, J.-O. and Kanandjembo, A. 2001. Comparison of horse mackerel length frequencies obtained from research vessels and commercial midwater trawlers: implications for biomass estimation. South African Journal of Marine Science 23: 265-274. Berry, H.U. 2002. Cape Cross: Past and present. The Authors, P&M Bridgeford, Walvis Bay, 44 p. Best, P.B. 2006. The presence of right whales in summer on the west coast of South Africa: the evidence from historical records. African Journal of Marine Science 28: 159-166. Best, P.B., Crawford, R.J.M. and Van der Elst, R.P. 1997. Top predators in Southern Africa's marine ecosystem. Transactions of the Royal Society of South Africa 52: 177-225. Beyers, C.J.D.B. and Wilke, C.G. 1980. Quantitative stock survey and some biological and morphometric characteristics of the deepsea red crab Geryon quinquedens off South West Africa. Fisheries Bulletin of South Africa 13: 9-19. Bianchi, G., Hamukuaya, H. and Alvheim, O. 2001. On the dynamics of demersal fish assemblages off Namibia in the 1990s. South African Journal of Marine Science 23: 419-428. Bianchi, G., Carpenter, K.E., Roux, J.-P., Molloy, F.J., Boyer, D. and Boyer, H.J. 1993. The living marine resources of Namibia. FAO species identification field guide for fishery purposes. FAO, NORAD, Rome, 250+ p. Booth, A.J. and Quinn, T.J.I. 2006. Maximum likelihood and Bayesian approaches to stock assessment when data are questionable. Fisheries Research 80: 169-181. Boyer, D. and Boyer, H. 2004. The large volume small pelagic fisheries of the Southeast Atlantic: Angola, Namibia and South Africa. pp. 67-94 in Swan, J. and Gréboval, D., editors. Report of the International workshop on the Implementation of International  64  Northern Benguela Ecosystem, Heymans and Sumaila Fisheries Instruments and Factors of Unsustainability and Overexploitation in Fisheries, Mauritius, 3-7 February 2003. FAO Fisheries Report No. 700. FAO, Rome.  Boyer, D.C. and Hampton, I. 2001. An overview of the living marine resources of Namibia. South African Journal of Marine Science 23(A Decade of Namibian Fisheries Science): 5-35. Boyer, D.C. and Oelofsen, B. 2004. Co-Management: Namibia's experience with two large-scale industrial fisheries - sardine and orange-roughy. pp. 333-356 in Sumaila, U.R., Steinshamn, S.I., Skog, M. and Boyer, D.C., editors. Ecological, economic and social aspects of Namibian fisheries. Eburon, Netherlands. Boyer, D.C., Boyer, H.J., Fossen, I. and Kreiner, A. 2001a. Changes in abundance of the northern Benguela sardine stock during the decade 1990-2000, with comments on the relative importance of fishing and the environment. South African Journal of Marine Science 23: 67-84. Boyer, D.C., Kirchner, C.H., McAllister, M.K., Staby, A. and Staalesen, B.I. 2001b. The orange roughy fishery of Namibia: lessons to be learned about managing a developing fishery. South African Journal of Marine Science 23: 205-221. Brown, P.C., Painting, S.J. and Cochrane, K.L. 1991. Estimates of phytoplankton and bacterial biomass and production in the northern and southern Benguela ecosystems. South African Journal of Marine Science 11: 537-564. Burmeister, L. 2005. Is there a single stock of Merluccius paradoxus in the Benguela ecosystem? African Journal of Marine Science 27: 23-32. Butterworth, D.S. and Geromont, H.F. 2001. Evaluation of a class of possible simple interim management procedures for the Namibian hake fishery. South African Journal of Marine Science 23: 357-374. Butterworth, D.S. and Rademeyer, R.A. 2005. Sustainable management initiatives for the southern African hake fisheries over recent years. Bulletin of Marine Science 76: 287-319. Butterworth, D.S., Hughes, G. and Strumpfer, F. 1990. VPA with ad hoc tuning: implementation for disaggregated fleet data, variance estimation, and application to the Namibian stock of Cape horse mackerel Tracurus trachurus capensis. South African Journal of Marine Science 9: 327-357. Butterworth, D.S., Punt, A.E., Oosthuizen, W.H. and Wickens, P.A. 1995. The effects of future consumption by the Cape fur seal on catches and catch rates of the Cape hakes. 3. Modelling the dynamics of the Cape fur seal Arctocephalus pusillus pusillus. South African Journal of Marine Science 16: 161-183. Carr, M.-E. 2002. Estimation of potential productivity in Eastern Boundary Currents using remote sensing. Deep-Sea Research II 49: 59-80. Carr, M.-E. and Kearns, E.J. 2003. Production regimes in four Eastern Boundary Current systems. Deep-Sea Research II 50: 31993221. Crawford, R.J.M. and De Villiers, G. 1985. Snoek and their prey-interrelationships in the Benguela upwelling system. South African Journal of Science 81: 91-97. Crawford, R.J.M., Shannon, L.V. and Pollock, D.E. 1987. The Benguela ecosystem. Part IV. The major fish and invertebrate resources. Oceanography and Marine Biology Annual Review 25: 353-505. Crawford, R.J.M., Shannon, L.V. and Shelton, P.A. 1989. Characteristics and management of the Benguela as a large marine ecosystem. pp. 169-219 in Sherman, K. and Alexander, L.M., editors. Biomass yields and geography of large marine ecosystems, American Association for the Advancement of Science Symposium 111. Westview Press, Inc., Boulder, Colorado. Crawford, R.J.M., Ryan, P.G. and Williams, A.J. 1991. Seabird consumption and production in the Benguela and Western Agulhas ecosystems. South African Journal of Marine Science 11: 357-375. Crawford, R.J.M., Cruickshank, R.A., Shelton, P.A. and Kruger, I. 1985. Partitioning of a goby resource amongst four avian predators and evidence for altered trophic flow in the pelagic community of an intense perennial upwelling system. South African Journal of Marine Science 3: 215-228. Crawford, R.J.M., David, J.H.M., Shannon, L.J., Kemper, J., Klages, N.T.W., Roux, J.-P., Underhill, L.G., Ward, V.L., Williams, A.J. and Wolfaardt, A.C. 2001. African penguins as predators and prey - coping (or not) with change. South African Journal of Marine Science 23: 435-447. Critchley, A.T., Rotmann, K.W.G. and Molloy, F.J. 1991. The Namibian seaweed industry: Present and potential. Bioresource Technology 38: 137-143. Cruickshank, R.A. 1983. Lanternfish ecology in the Benguela current system. South African Journal of Science 79: 149-150. Cruickshank, R.A., Cooper, J. and Hampton, I. 1980. Extension to the geographical distribution of pelagic goby Sufflogobius bibarbatus off South West Africa and some mensural and energetic information. Fisheries Bulletin of South Africa 13: 77-82. Cury, P. and Shannon, L.J. 2004. Regime shifts in upwelling ecosystems: observed changes and possible mechanisms in the northern and southern Benguela. Progress in Oceanography 60: 223-243. Cushing, D.H., Humphrey, G.F., Banse, K. and Laevastu, T. 1958. Report of the Committee on Terms and Equivalents. Rapports et Procès-verbaux des Réunions du Conseil Permanent International pour l'Exploration de la Mer 14: 15-16. David, J.H.M. 1987. Diet of the South African fur seal (1974-1985) and an assessment of competition with fisheries in southern Africa. South African Journal of Marine Science 5: 693-713. David, J.H.M. 1989. Seals. pp. 288-302 in Payne, A.I.L. and Crawford, R.J.M., editors. Oceans of life off Southern Africa. Vlaeberg Publishers, Cape Town.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  65  De Villiers, D.J., Oosthuizen, W.H., Roux, J.-P. and Kotze, P.G.H. 1997. Population structure of a non-breeding colony of the Cape fur seal Arctocephalus pusillus pusillus at Cape Frio, Namibia. South African Journal of Marine Science 18: 295-298. Ebert, D.A. 1996. Biology of the sevengill shark Notorynchus cepedianus (Peron, 1807) in the temperate coastal waters of Southern Africa. South African Journal of Marine Science 17: 93-103. Fearon, J.J., Boyd, A.J. and Schülein, F.H. 1992. Views on the biomass and distribution of Chrysaora hysoscella (Linne, 1766) and Aequorea aequorea (Forskal, 1775) off Namibia, 1982-1989. Scientia Marina 56(1): 75-85. Findlay, K.P., Best, P.B., Ross, G.J.B. and Cockcroft, V.G. 1992. The distribution of small odontocete cetaceans off the coasts of South Africa and Namibia. South African Journal of Marine Science 12: 237-270. Fossen, I., Boyer, D.C. and Plarre, H. 2001. Changes in some key biological parameters of the northern Benguela sardine stock. South African Journal of Marine Science 23: 111-121. Froese, R., Pauly, D. 2000. FishBase 2000: concepts, design and data sources. [Distributed with 4 CD-ROMs]; www.fishbase.org, 346 pp. in. ICLARM, Los Baños, Philippines. Fuller, B. and Prommer, I. 2000. Population-Development-Environment in Namibia. Background readings. Interim Report IR-0031, International Institute for Applied Systems Analysis, Laxenburg, Austria, 326 p. Griffiths, C.L., Van Sittert, L., Best, P.B., Brown, A.C., Clark, B.M., Cook, P.A., Crawford, R.J.M., David, J.H.M., Davies, B.R., Griffiths, M.H., Hutchings, K., Jerardino, A., Kruger, N., Lamberth, S., Leslie, R.W., Mellville-Smith, R., Tarr, R. and Van der Lingen, C.D. 2004. Impacts of human activities on marine animal life in the Benguela: A historical overview. Oceanography and Marine Biology Annual Review 42: 303-392. Griffiths, M.H. 2002. Life history of South African snoek, Thyrsites atun (Pices: Gempylidae): a pelagic predator of the Benguela ecosystem. Fisheries Bulletin 100: 690-710. Griffiths, M.H. 2003. Stock structure of snoek Thyrsites atun in the Benguela: a new hypothesis. African Journal of Marine Science 25: 383-386. Grobler, C.A.F. and Noli-Peard, K.R. 1997. Jasus lalandii fishery in post-independence Namibia: monitoring population trends and stock recovery in relation to a variable environment. Marine and Freshwater Research 48: 1015-1022. Hampton, I. 2003. Harvesting the sea. pp. 43-80 in Anon., editor. A marine state of Environment Report: Namibia's Marine Environment. Namibian Ministry of Environment and Tourism. Hamukuaya, H., O'Toole, M.J. and Woodhead, P.M.J. 1998. Observations of severe hypoxia and offshore displacement of Cape hake over the Namibian shelf in 1994. South African Journal of Marine Science 19: 57-59. Hamukuaya, H., Bianchi, G. and Baird, D. 2001. The structure of demersal assemblages off Namibia in relation to abiotic factors. South African Journal of Marine Science 23: 397-417. Hansen, F.C., Cloete, R.R. and Verheye, H.M. 2005. Seasonal and spatial variability of dominant copepods along a transect off Walvis Bay (23°S), Namibia. African Journal of Marine Science 27: 55-63. Hart, T.J. and Currie, R.I. 1960. The Benguela Current. Discovery Report 31: 123-298. Hernández, I., Fernández-Engo, M.A., Pérez-Lloréns, J.L. and Vergara, J.J. 2005. Integrated outdoor culture of two estuarine macroalgae as biofilters for dissolved nutrients from Sparus aurata waste waters. Journal of Applied Phycology 17: 557-567. Hewitson, J.D. and Cruickshank, R.A. 1993. Production and consumption by planktivorous fish in the Northern Benguela Ecosystem in the 1980s. South African Journal of Marine Science 13: 15-24. Heymans, J.J. 1997. Network analysis of the carbon flow model of the northern Benguela ecosystem, Namibia. PhD. University of Port Elizabeth, now Nelson Mandela Metropolitan University, Port Elizabeth Pages p. Heymans, J.J. 2004. The effects of internal and external control on the northern Benguela ecosystem. pp. 29-52 in Sumaila, U.R., Steinshamn, S.I., Skog, M. and Boyer, D.C., editors. Ecological, economic and social aspects of Namibian fisheries. Eburon, Netherlands. Heymans, J.J. and Baird, D. 2000. Network analysis of the northern Benguela ecosystem by means of NETWRK and ECOPATH. Ecological Modelling 131: 97-119. Holtzhausen, J.A. 1999. Population dynamics and life history of westcoast steenbras (Lithognathus aureti (Sparidae)), and management options for the sustainable exploitation of the steenbras resource in Namibian waters. PhD. University of Port Elizabeth, Port Elizabeth Pages p. Holtzhausen, J.A. and Kirchner, C.H. 2004. Management regulations of Namibian angling fish species. pp. 113-134 in Sumaila, U.R., Steinshamn, S.I., Skog, M. and Boyer, D.C., editors. Ecological, economic and social aspects of Namibian fisheries. Eburon, Netherlands. Holtzhausen, J.A., Kirchner, C.H. and Voges, E. 2001. Observations on the linefish resources of Namibia, 1990-2000, with special reference to west coast steenbras and silver kob. South African Journal of Marine Science 23: 135-144. Hugo, F. 2006. Sea Shepherd News Release: Seal Alert appeals to CITES to take action for South African seals. pp. Letter to CITES in. Sea Shepherd Conservation Society. Huse, I., Hamukuaya, H., Boyer, D.C., Malan, P.E. and Strømme, T. 1998. The diurnal vertical dynamics of Cape hake and their potential prey. South African Journal of Marine Science 19: 365-376.  66  Northern Benguela Ecosystem, Heymans and Sumaila  Hutchings, L., Pillar, S.C. and Verheye, H.M. 1991. Estimates of standing stock, production and consumption of meso- and macrozooplankton in the Benguela ecosystem. South African Journal of Marine Science 11: 499-512. Jarre-Teichmann, A., Shannon, L.J., Moloney, C.L. and Wickens, P.A. 1998. Comparing trophic flows in the southern Benguela to those in other upwelling ecosystems. South African Journal of Marine Science 19: 391-414. Kemper, J., Roux, J.-P., Bartlett, P.A., Chesselet, Y.J., James, J.A.C., Jones, R., Wepener, S. and Molloy, F.J. 2001. Recent population trends of African penguins Spheniscus demersus in Namibia. South African Journal of Marine Science 23: 429-434. Kinahan, J. 1991. The historical archaeology of nineteenth century fisheries at Sandwich Harbour on the Namib coast. Cimbebasia 13: 1-27. Kirchner, C.H., Sakko, A.L. and Barnes, J.I. 2000. An economic valuation of the Namibian recreational shore-angling fishery. South African Journal of Marine Science 22: 17-25. Klingelhoeffer, E.W. 2006. Population dynamics and the development of a sustainable fishery for the Cape horse mackerel Trachurus capensis (Castelnau, 1861), in the northern Benguela current upwelling system. PhD. Nelson Mandela Metropolitan University, Port Elizabeth Pages p. Konchina, Y.V. 1986. Distribution and feeding of the South African horse mackerel and hake in the Namibian shelf waters. Collection of Scientific Papers of the international Commision of South East Atlantic Fisheries 13(2): 7-18. Krakstad, J.-O. and Kanandjembo, A. 2001. A summary of the fishery and biology of Cape horse mackerel (Trachurus capensis) in Namibian waters. pp. 1-15 in Payne, A.I.L., editor. Proceedings of an international workshop on the research and managment of horse mackerel in Namibian waters. Ministry of Fisheries and Manrine Resources, Swakopmund. Kreiner, A., Van der Lingen, C.D. and Fréon, P. 2001. A comparison of condition factor and gonadosomatic index of sardine Sardinops sagax stocks in the northern and southern Benguela upwelling ecosystems, 1984-1999. South African Journal of Marine Science 23: 123-134. le Clus, F. 1990. Impact and implications of large-scale environmental anomalies on the spatial distribution of spawning of the Namibian pilchard and anchovy populations. South African Journal of Marine Science 9: 141-159. le Clus, F., Melo, Y.C. and Cooper, R.M. 1988. Impact of environmental perturbations during 1986 on the availability and abundance of pilchard and anchovy in the northern Benguela system. Collection of Scientific Papers of the international Commision of South East Atlantic Fisheries 15(2): 49-70. Le Roux, L. 2001. The impact of emigration on population estimates of deep-sea red crab Chaceon maritae off Namibia. African Journal of Marine Science 23: 61-66. Lees, R. 1969. Fishing for fortunes: The story of the fishing industry in Southern Africa - and the men who made it. Purnell, Cape Town, 283 p. I only have some pages, not the whole document. Lipiński, M.R. 1992. Cephalopods and the Benguela ecosystem: Trophic relationships and impact. South African Journal of Marine Science 12: 791-802. Lynam, C.P., Gibbons, M.J., Axelsen, B.E., Sparks, C.A.J., Coetzee, J., Heywood, B.G. and Brierley, A.S. 2006. Jellyfish overtake fish in an heavily fished ecosystem. Current Biology 16: 492-493. Maartens, L. and Booth, A.J. 2001. Quantifying commercial catch and effort of monkfish Lophius vomerinus and L. vaillanti off Namibia. South African Journal of Marine Science 23: 291-306. Maartens, L. and Booth, A.J. 2005. Aspects of the reproductive biology of monkfish Lophius vomerinus off Namibia. African Journal of Marine Science 27: 325-329. Macpherson, E. and Gordoa, A. 1992. Trends in the demersal fish community off Namibia from 1983 to 1990. South African Journal of Marine Science 12: 635-649. Maurihungirire, M. 2004. Transboundary issues in the purse-seine, trawl and crustacean fisheries of the Southeast Atlantic. pp. 151164 in Payne, A.I.L., O'Brien, C.M. and Rogers, S.I., editors. Management of shared fish stocks. Blackwell Publishers, Oxford. Mecenero, S., Kirkman, S.P. and Roux, J.-P. 2005. Seabirds in the diet of Cape fur seals Arctocephalus pusillus pusillus at three mainland breeding colonies in Namibia. African Journal of Marine Science 27: 509-512. Mecenero, S., Roux, J.-P., Underhill, L.G. and Bester, M.N. 2006. Diet of Cape fur seals Arctocephalus pusillus pusillus at three mainland breeding colonies in Namibia. 1. Spatial variation. African Journal of Marine Science 28: 57-71. Melo, Y.C. and le Clus, F. 2005. Growth and reproduction of the pelagic goby Sufflogobius bibarbatus off the Orange River, southern Africa. African Journal of Marine Science 27: 265-273. Melville-Smith, R. 1983. Some observations on the abundance of deep-sea red crab Geryon maritae in Namibian waters using photography. South African Journal of Science 79: 144. Melville-Smith, R. 1988. The commercial fishery for and population dynamics of red crab Geryon maritae off South West Africa, 1976-1986. South African Journal of Marine Science 6: 79-75. Miller, C.M., Oosthuizen, W.H. and Wickens, P.A. 1996. Cape fur seals trapped in trawling gear: age structure, sex ratio, seasonality and distribution. South African Journal of Marine Science 17: 105-111. Nicols, P. 2004. Marine fisheries management in Namibia: Has it worked? pp. 319-332 in Sumaila, U.R., Steinshamn, S.I., Skog, M. and Boyer, D.C., editors. Ecological, economic and social aspects of Namibian fisheries. Eburon, Netherlands.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  67  Pagès, F. 1992. Mesoscale coupling between planktonic Cnidarian distribution and water masses during a temporal transition between active upwelling and abatement in the northern Benguela system. South African Journal of Marine Science 12: 41-52. Pillar, S.C. and Barange, M. 1997. Diel variability in bottom trawl catches and feeding activity of the Cape hakes off the west coast of South Africa. ICES Journal of Marine Science 54: 485-499. Pitcher, G.C., Brown, P.C. and Mitchell-Innes, B.A. 1992. Spatio-temporal variability of phytoplankton in the southern Benguela upwelling system. South African Journal of Marine Science 12: 439-456. Pollock, D.E. and Shannon, L.V. 1987. Response of rock-lobster populations in the Benguela ecosystem to environmental change - a hypothesis. South African Journal of Marine Science 5: 887-899. Pollock, D.E., Cockcroft, A.C., Groeneveld, J.C. and Schoeman, D.S. 2000. Chapter 5: The commercial fisheries for Jasus and Palinurus species in the south-east Atlantic and south-west Indian Oceans. pp. 105-120 in Phillips, B.F. and Kittaka, J., editors. Spiny lobsters: Fisheries and culture. Blackwell Science Ltd., Oxford. Pulfrich, A., Parkins, C.A. and Branch, G.M. 2003. The effects of shore-based diamond-diving on intertidal and subtidal biological communities and rock lobsters in southern Namibia. Aquatic Conservation: Marine and Freshwater Ecosystems 13: 233-255. Punt, A.E., Leslie, R.W. and Du Plessis, S.E. 1992. Estimation of the annual consumption of food by Cape hake Merluccius capensis and M. paradoxus of the South African west coast. South African Journal of Marine Science 12: 611-634. Punt, A.E., Butterworth, D.S. and Penney, A.J. 1995. Stock assessment and risk analysis for the South Atlantic population of albacore Thunnus alalunga using an age-structured production model. South African Journal of Marine Science 16: 287-310. Roel, B.A. and Macpherson, E. 1988. Feeding of Merluccius capensis and M. paradoxus off Namibia. South African Journal of Marine Science 6: 227-243. Romanov, E. 2001. History of Ukranian fisheries in the South-eastern Atlantic and YugNIRO database for this area. in Payne, A.I.L., editor. Proceedings of an international workshop on the research and managment of horse mackerel in Namibian waters. Ministry of Fisheries and Manrine Resources, Swakopmund. Rotmann, K.W.G. 1987. The collection, utilization and potential farming of red seaweeds in Namibia. Hydrobiologia 151/152: 301305. Roux, J.-P. and Shannon, L.J. 2004. Ecosystem approach to fisheries management in the northern Benguela: the Namibian experience. African Journal of Marine Science 26: 79-93. Roux, J.-P., Best, P.B. and Stander, P.E. 2001. Sightings of southern right whales (Eubalaena australis) in Namibian waters, 19711999. Journal of Cetacean research and management Special Issue 2: 181-185. Roy, C., Cury, P. and Kifani, S. 1992. Pelagic fish recruitment success and reproductive strategy in upwelling areas: environmental compromises. South African Journal of Marine Science 12: 135-146. Ryan, P.G., Keith, D.G. and Kroese, M. 2002. Seabird bycatch by tuna longline fisheries off Southern Africa, 1998-2000. South African Journal of Marine Science 24: 103-110. Ryther, J.H. 1969. Photosynthesis and fish production in the sea. Science, New Series 166: 72-76. Schwartzlose, R.A., Alheit, J., Bakun, A., Baugartner, T.R., Cloete, R., Crawford, R.J.M., Fletcher, W.J., Green-Ruiz, Y., Hagen, E., Kawasaki, T., Lluch-Belda, D., Lluch-Cota, S.E., MacCall, A.D., Matsuura, Y., Nevárez-Martínez, M.O., Parrish, R.H., Roy, C., Serra, R., Shust, K.V., Ward, M.N. and Zuzunaga, J.Z. 1999. Worldwide large-scale fluctuations of sardine and anchovy populations. South African Journal of Marine Science 21: 289-347. Shannon, L.J. and Taunton-Clark, J. 1988. Interannual and decadal changes in sea surface temperature and relative wind stress in the South-East Atlantic this century. pp. 49-51 in MacDonald, I.A.W. and Crawford, R.J.M., editors. Long-term data series relating to southern African's renewable natural resources. Foundation for Research Development, Pretoria. Shannon, L.J. and Jarre-Teichmann, A. 1999. A model of trophic flows in the Northern Benguela upwelling system during the 1980s. South African Journal of Marine Science 21: 349-366. Shannon, L.V. 1985. The Benguela ecosystem. Part I. Evolution of the Benguela, physical features and processes. Oceanography and Marine Biology Annual Review 23: 105-182. Shannon, L.V. and Pillar, S.C. 1986. The Benguela ecosystem. Part III. Plankton. Oceanography and Marine Biology Annual Review 24: 65-170. Shannon, L.V. and O'Toole, M.J. 2003. Sustainability of the Benguela ex Africa semper aliquid novi. pp. 227-253 in Hempel, G. and Sherman, K., editors. Large Marine Ecosystems of the World: Trends in Exploitation, Protection and Research. Elsevier. Shannon, L.V., Boyd, A.J., Brundrit, G.B. and Taunton-Clark, J. 1986. On the existence of an El Niño-type phenomenon in the Benguela system. Journal of marine Research 44(3): 495-520. Shelton, P.A. and Crawford, R.J.M. 1988. Analysis of time series of seabird guano harvested off southern Africa. pp. 51-53 in MacDonald, I.A.W. and Crawford, R.J.M., editors. Long-term data series relating to southern African's renewable natural resources. Foundation for Research Development, Pretoria. Shin, Y.-J., Shannon, L.J. and Cury, P.M. 2004. Simulations of fishing effects on the southern Benguela fish community using an individual-based model: learning from a comparison with Ecosim. African Journal of Marine Science 26: 95-114. Sparks, C., Buecher, E., Brierley, A.S., Axelsen, B.E., Boyer, H. and Gibbons, M.J. 2001. Observations on the distribution and relative abundance of the scyphomedusan Chrysaora hysoscella (Linné, 1766) and the hydrozoan Aequorea aequorea (Forskål, 1775) in the northern Benguela ecosystem. Hydrobiologia 451: 275-286.  68  Northern Benguela Ecosystem, Heymans and Sumaila  Sparks, D.L. 1984. Namibia's coastal and marine development potential. African Affairs 83: 477-496. Stage, J. and Kirchner, C.H. 2005. An economic comparison of the commercial and recreational linefisheries in Namibia. African Journal of Marine Science 27: 577-584. Stenevik, E.K., Skogen, M.D., Sundby, S. and Boyer, D.C. 2003. The effect of vertical and horizontal distribution retention of sardine (Sardinops sagax) larvae in the Northern Benguela - observations and modeling. Fisheries Oceanography 12: 185-200. Sumaila, U.R., Ninnes, C. and Oelofsen, B. 2002. Management of shared hake stocks in the Benguela marine ecosystem. pp. 143-158 in FAO, editor. Papers presented at the Norway-FAO Expert Consultation on the Management of Shared Fish Stocks, Bergen, Norway 7-10 October 2002. FAO Fisheries Report No. 695. FAO, Rome. Sumaila, U.R., Marsden, A.D., Watson, R. and Pauly, D. 2007. Global ex-vessel fish price database: construction and applications. Journal of Bioeconomics, in press. Thomas, R.M. 1986. The Namibian Pilchard: The 1985 season, assessment for 1952-1985 and recommenations for 1986. ICSEAF Collectin of scientific Papers of the International Commision for Southeast Atlantic Fisheries 13: 245-269. Timonin, A.G., Araschkevich, E.G., Drits, A.V. and Semenova, T.N. 1992. Zooplankton dynamics in the northern Benguela ecosystem, with special reference to the copepod Calanoides carinatus. South African Journal of Marine Science 12: 545-560. Van der Westhuizen, A. 2001. A decade of exploitation and management of the Namibian hake stocks. South African Journal of Marine Science 23: 307-315. Vaske, B. 2001. Some comments on the biological effects of the small-meshed purse-seine fishery. in Payne, A.I.L., editor. Proceedings of an international workshop on the research and managment of horse mackerel in Namibian waters. Ministry of Fisheries and Marine Resources, Swakopmund. Venter, G.E. 1988a. Occurrence of jellyfish on the west coast of South West Africa/Namibia. pp. 54-61 in MacDonald, I.A.W. and Crawford, R.J.M., editors. Long-term data series relating to southern African's renewable natural resources. Foundation for Research Development, Pretoria. Venter, J.D. 1988b. Fluctuations in the catches of some linefish species off South West Africa/Namibia, 1964-1986. pp. 62-64 in MacDonald, I.A.W. and Crawford, R.J.M., editors. Long-term data series relating to southern African's renewable natural resources. Foundation for Research Development, Pretoria. Verheye, H.M., Hutchings, L., Huggett, J.A. and Painting, S.J. 1992. Mesozooplankton dynamics in the Benguela ecosystem, with emphasis on the herbivorous copepods. South African Journal of Marine Science 12: 561-584. Voges, E., Gordoa, A., Bartholomae, C.H. and Field, J.G. 2002. Estimating the probability of different levels of recruitment for Cape hakes Merluccius capensis off Namibia, using environmental indices. Fisheries Research 58: 333-340. Walmsley, S.A., Leslie, R.W. and Sauer, W.H.H. 2005. The biology and distribution of the monkfish Lophius vomerinus off South Africa. African Journal of Marine Science 27: 157-168. Wasmund, N., Lass, H.-U. and Nausch, G. 2005. Distribution of nutrients, chlorophyll and phytoplankton primary production in relation to hydrographic structures bordering the Benguela-Angola frontal region. African Journal of Marine Science 27: 177190. Wickens, P.A., David, J.H.M., Shelton, P.A. and Field, J.G. 1991. Trends in harvests and pup numbers of the South African fur seal: Implications for management. South African Journal of Marine Science 11: 307-326. Willemse, N.E. 2002. Major trends in the marine fisheries catches of Namibia, 1950-2000. Norwegian College of Fisheries Science, University of Tromsø, Tromsø. Zeybrandt, F. and Barnes, J.I. 2001. Economic characteristics of demand in Namibia's marine recreational shore fishery. South African Journal of Marine Science 23: 145-156.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  69  Appendix A: Diet matrix of the northern Benguela model in 1956. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33  Group Marine mammals Seals Birds Sharks Tuna Snoek Other linefish Anchovy juveniles Anchovy adults Sardine juveniles Sardine adults Gobies juveniles Gobies adults Other s pelagics Mesopelagics Juv. h mackerel Adult h mackerel J hake A hake Monkfish Other demersals Cephalopods Macrobenthos Crabs Lobster Jellyfish juvs Jellyfish adults Macrozooplankton Mesozooplankton Benthic producers Phytoplankton Detritus Import  1  2  3  0.00008  0.00001 0.001  0.026 0.0001  0.003  0.5  0.253  4 0.001 0.001  5  6  7  0.091  0.05  0.562  8  9  10  11  12  0.093  0.101  0.011  0.069 0.117 0.193  0.429 0.037 0.053 0.005  0.1 0.175  0.059 0.106 0.007 0.1 0.036 0.01 0.025  0.318  0.03  0.005  0.012  0.091  0.175  0.11  0.012 0.097 0.042  0.005 0.145 0.05 0.01 0.044 0.003 0.099  0.0001  0.007 0.0001  0.012 0.01  0.108 0.044  0.043 0.003  15  0.0001  0.001 0.014  14  0.144 0.001 0.002 0.002 0.0001  0.014  13  0.004 0.05 0.365 0.1 0.054 0.05  0.05 0.006 0.024  0.091  0.121  0.01 0.007 0.09 0.01 0.026  0.165 0.111  0.012  0.003 0.01  0.036 0.055  0.041 0.033 0.014  0.254  1 0.375  0.0001  0.00001  0.00001  0.00001  0.32 0.31  0.18 0.18  0.06 0.01  0.33 0.038  1  0.56 0.08  1  0.93  0.021 0.35 0.597  0.6 0.4  70  Appendix A continued: Diet matrix of the northern Benguela model in 1956. Group 16 17 18 19 20 21 1 Marine mammals 2 Seals 3 Birds 4 Sharks 5 Tuna 6 Snoek 0.001 7 Other linefish 0.0001 8 Anchovy juveniles 9 Anchovy adults 0.018 0.01 0.039 10 Sardine juveniles 11 Sardine adults 0.018 0.045 12 Gobies juveniles 13 Gobies adults 0.009 0.018 0.045 0.102 0.021 14 Other s pelagics 0.045 0.197 0.054 15 Mesopelagics 0.02 0.01 0.01 0.129 0.018 16 Juv. h mackerel 0.01 0.066 17 Adult h mackerel 0.01 0.05 18 J hake 0.005 0.001 0.00099 19 A hake 0.05 0.26 0.009 20 Monkfish 0.0005 0.01 0.00094 21 Other demersals 0.005 0.05 0.009 22 Cephalopods 0.001 0.014 0.005 0.034 0.00094 23 Macrobenthos 0.018 0.045 0.062 24 Crabs 0.009 25 Lobster 0.009 26 Jellyfish juvs 27 Jellyfish adults 0.045 28 Macrozooplankton 0.79 0.816 0.311 0.304 29 Mesozooplankton 1 0.18 0.088 0.358 0.11 30 Benthic producers 31 Phytoplankton 32 Detritus 0.455 33 Import  Northern Benguela Ecosystem, Heymans and Sumaila  22  23  24  25  26  27  28  29  0.001 0.001  0.1 0.1 0.1 0.01 0.001 0.01 0.01 0.01 0.05 0.001 0.001  0.333  0.5  0.394 0.285 0.05  1 0.898  0.25  0.4  0.247 0.5  0.6  0.5 0.647  0.5 0.5  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  71  MODELLING THE FOOD WEB IN THE UPWELLING ECOSYSTEM OFF CENTRAL CHILE (33°S–39°S) IN THE YEAR 20001 Sergio Neira  Zoology Department, University of Cape Town, Rondebosh 7701, Cape Town, South Africa; Email: sergio.neiraalarcon@uct.ac.za  Hugo Arancibia  Departamento de Oceanografía, Universidad de Concepción, Casilla 160-C, Concepción, Chile  ABSTRACT We describe the main steps in the construction of a food web model representing trophic relationships for the main fishery resources in the upwelling system off central Chile. We use biological and fishing data for the most important functional groups covering the year 2000 and the Ecopath with Ecosim software. The model includes 31 functional groups, from primary producers to top predators such as marine birds and mammals. This paper presents a comprehensive description of the study area, the functional groups included and the source of the data needed to parameterize the model. We also present key assumptions in the modelling process derived from the lack of information on basic life history parameters for functional groups.  INTRODUCTION Marine populations are tightly linked to their physical and biological environment. This basic ecological principle is normally disregarded by traditional single-species fisheries management, by considering stocks as discrete and independent units (Walters and Martel, 2004). A side effect of this assumption is that fishing mortality was considered to be the only factor explaining stock dynamics. This myopic approach cannot assess or predict the effects of environmental and trophic (predator and prey) controls on marine populations, which are as important as fishing in explaining stock fluctuations (Cushing, 1982; Sinclair et al., 1997; Bax, 1991, 1998). Therefore, it is likely that ignoring interactions between stocks and the ecosystem, from where landings are removed, played a significant role in the worldwide overexploitation crisis of fisheries (Pauly et al., 2003). The global call for an ecosystem approach to fisheries (EAF) implies that future stock assessment and management should not rely exclusively on single species approaches (FAO, 2003). In addition, EAF pre-supposes an incremental knowledge on the key aspects of the ‘ecology of the ecosystem’, which is expected to play an important role in future stock recovery and fisheries sustainability (Botsford et al., 1997; Link, 2002a, b; Pikitch et al., 2004). Changes in predator-prey interactions and energy/mass flows among trophic levels are important aspects of ecosystem structure (Elton, 1927; Lindeman, 1942). An increasing body of literature indicates that these two ecosystem properties are negatively altered by fishing (Pauly and Christensen, 1995; Pauly et al., 1998a; Worm et al., 2006). Therefore, the understanding of ecosystem trophodynamics should be an important aspect of EAF. The development of trophic interaction, food web and ecosystem models that appear capable of making useful predictions about ecosystem-based policy issues, ignored in single-species assessment and policy Cite as: Neira, S. and Arancibia, H. 2007. Modelling the food web in the upwelling ecosystem off central Chile (33°S–39°S) in the year 2000, p. 71–86. In: Le Quesne, W.J.F., Arreguín-Sánchez, F. and Heymans, S.J.J. (eds.) INCOFISH ecosystem models: transiting from Ecopath to Ecospace. Fisheries Centre Research Reports 15(6). Fisheries Centre, University of British Columbia [ISSN 1198-6727]. 1  72  Upwelling Ecosystem off Central Chile, Neira and Arancibia  recommendations, are regarded as potential improvement for fisheries stock assessment and management (Walters and Martell, 2004). The Ecopath with Ecosim model and software (EwE; Christensen and Pauly, 1992; Walters et al., 1997; Pauly et al., 2000) represents a well-known approach to analyse predator-prey relationships and energy flows (i.e., who eats whom? and how much?) in aquatic ecosystems. EwE’s friendly interface and easily parameterized framework has resulted in dozens of published models describing aquatic ecosystems all over the globe (Christensen and Pauly, 1993; Christensen and McLean, 2004; www.ecopath.org). Previous representations of the food web in the upwelling system off central Chile have been published (Neira and Arancibia, 2004) describing system conditions in the early 1990s, i.e., healthy stocks and no severe changes in environmental conditions. By the mid- to late 1990s, however, the system experienced changes in fishing and environmental forcing. The most important changes were related to the ENSO event in 1997–1998, the strongest of the 20th century (MacPhaden, 1999), and sequential stock collapses of horse mackerel Trachurus symmetricus (1997), red squat lobster Pleuroncodes monodon (2000) and hake Merluccius gayi (2004). In the late 1990s, the dominant species of small pelagic fish, common sardine (Strangomera bentincki) and anchovy (Engraulis ringens), suffered heavy fishing pressure. In addition, unusual invasions of jumbo squid (Dosidicus gigas) have been reported since the early 2000s (Arancibia et al., 2007). It is likely that these changes affected energy flows and prey-predator relationships through the food web resulting in an altered community structure. In this paper we describe an Ecopath model constructed to analyze food web structure off central Chile at the beginning of the 2000s. We consider that this model will allow future comparisons with both previous models constructed for the Chilean system (i.e., Neira and Arancibia, 2004) and other upwelling systems. This report describes the data sources and calculations used to parameterize an Ecopath model of the upwelling system off central Chile (USCCh).  METHODS The USCCh model includes biological and fishing information of the main functional groups inhabiting the study area, with emphasis on fishing resources, their prey and predators in the year 2000. The food web model was constructed using EwE version 5.1. Below we present a comprehensive description of the study area, the functional groups included and the source of the data needed to parameterize the model. We also present key assumptions in the modelling process derived from the lack of information for selected functional groups.  Study area The study area corresponds to the upwelling system off central Chile (USCCh; Figure 1) that is located in the southern section of the Humboldt Current System, which is one of the four major eastern boundary ecosystems of the world. The USCCh supports one of the highest levels of primary productivity recorded for the open ocean (19.9 g C m-2 d-1; Daneri et al., 2000) and globally significant landings (>4.5 million tons in 1995). Despite this high biological productivity, the state of exploited stocks is far from healthy (Arancibia and Neira, 2003). In fact, the main target species have been fully exploited or even overexploited for many years, leading to a series of recent stock collapses, for example of horse mackerel (1998), red squat lobster and yellow squat lobster (1999) and hake (2004). The USCCh, as considered in this modelling exercise, extends from 33ºS to 39ºS and from the coastline up to 30 nautical miles westward, covering a total area of approximately 50,000 km2. This geographical unit corresponds to the ‘Mediterranean District’ and is ecologically independent from the ‘Peruvian Province’ and the ‘Austral District’ located northward and southward, respectively (Camus, 2001). The main oceanographic and biogeographic patterns that characterize the study area are a rather narrow continental shelf (<30 nautical miles), strongly seasonal upwelling (September to March) and high levels of primary productivity (Strub et al., 1998; Daneri et al., 2000; Escribano et al., 2003). Oceanographically, four main water masses are present in the USCCh: Subtropical Surface Water (STSW), Subantarctic Water (SAW), Antarctic Intermediate Water (AAIW), and Equatorial Subsurface Water (ESSW) (for a review of the physical and chemical characteristics of each water mass, see Strub et al.,  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  73  1998). Wind-driven coastal upwelling brings the ESSW to the surface in the coastal zone, causing a persistent and characteristic shallow oxygen minimum zone (>0.5 ml O2.l-1). Based on the correlation between the low frequency coastal modes and the Pacific and Atlantic sea surface temperature (SST), Montecinos et al. (2003) suggest that the coastal SST comprises two main large-scale climate processes affecting the study area. At the inter-annual scale, the main source of variability is the El Niño Southern Oscillation cycle. At the long-term scale, an inter-decadal oscillation occurs at a basin-wide, and maybe even global, scale. In terms of the main biological components, the USCCh sustains a diverse and productive food web. The phytoplankton group is dominated by large diatoms for most of the year (Avaria and Muñoz, 1982), while the zooplankton is dominated by herbivorous copepods and euphausiids. Jellyfish (Hydrozoa) also constitute an important group in the plankton domain (Palma and Rosales, 1995). Macrocrustaceans are significant benthic components and some species such as red squat lobster (Pleuroncodes monodon), yellow squat lobster (Cervimunida johni) and pink shrimp (Heterocarpus reedi) support important fisheries. The fish community is dominated by pelagic species. Small pelagic fishes such anchovy (E. ringens) and the endemic common sardine (S. bentincki) are present at high biomass Figure 1. Study area corresponding to the upwelling system off levels off central Chile and dominate central Chile (highlighted zone) extending from 33º to 39ºS and landings. These species seem to feed from the line coast up to 30 nautical miles westward. primarily on phytoplankton and secondarily on zooplankton (Arrizaga et al., 1993). The horse mackerel (T. symmetricus) is a major fishery resource in the study area. This highly migratory species performs large-scale migrations in the Pacific Ocean. Off Chile, horse mackerel feed mainly on euphausiids (Miranda et al., 1998). The demersal fish community is dominated by hake (M. gayi) both in biomass and landings. Hake inhabits mid-depth waters (100–400 m) and feeds on euphausiids, galatheid crustaceans and small pelagic fish (Meléndez, 1983; Arancibia, 1989; Cubillos et al., 2003) and is highly cannibalistic (Arancibia et al., 1998). The USCCh also represents an independent management unit, comprising the main fishing ground for the Chilean purse seine and trawling fleets, both industrial and small scale operations and accounting for approximately 75% of the total landings in Chile (Neira and Arancibia, 2004; Neira et al., 2004). The industrial fishery, based on fish and crustacean species, started in the 1940s, when demersal trawlers targeted hake. However, landings of this fleet were significant only from the mid-1950s onwards. By the early 1960s, an industrial pelagic fishery came into operation, targeting small pelagic fishes, mainly common sardine and anchovy. At the same time, an industrial fleet operated on a medium-sized pelagic fish, namely horse mackerel, with landings of this species becoming significant from 1975 onwards.  74  Upwelling Ecosystem off Central Chile, Neira and Arancibia  Description of functional groups In this section we describe the main taxonomic groups covered using the letters A–G and the functional groups (in italics) included in the model representing the food web in the upwelling system off central Chile for the year 2000. The data sources for the input parameters are presented in Table 1. A. Top predators Top predators are important in food webs since they affect the abundance of other species either directly (by preying on them or competing with them) or indirectly (by removing other predators). Because of that, changes in higher trophic levels result in cascading effects that can spread through the entire food web (Frank et al., 2005). Top predators are also important for conservation policies since many of them (i.e., sharks, marine mammals and marine birds) are considered charismatic groups that are indirectly affected by overfishing of their prey. In spite of accurate data, we considered three ‘generic’ groups as top predators in the system: cetaceans, sea lions and marine birds. Cetaceans Taking into account the spatial scale of the model (30 nautical miles to the west) we only considered species with coastal distribution. Following Aguayo-Lobo et al. (1998), this functional group includes the dolphins (Cephalorhyncus eutropia Gray, 1846) and killer whales (Orcinus orca Linnaeus, 1758). No estimates of abundance, biomass, DC, P/B nor Q/B exist for this species in central Chile. Therefore, we use values for these parameters informed for similar species in comparable ecosystems. Sea lions The common sea lion (Otaria flavescens Shaw, 1800) is by far the most conspicuous species inhabiting the Chilean coast (Siefeld, 1999). Time series data of catches and abundance for this species in the study area are scarce. The first records indicate an intensive activity during the early 1900s. For example, in only one site (Santa Maria Island, 37º0.3’S; 73º31’W), 52,000 individuals were caught between 1921 and 1922 (Osgood, 1943). Currently, the common sea lion population in the study area is estimated at only 17,000 individuals (Doppler Ltda., 1997). Marine birds The functional group labelled ‘marine birds’ includes species such as seagulls (Larus dominicanus), the Humboldt penguin (Spheniscus humboldtii), pelicans (Pelicanus tagus) and cormorants (Phacalocorax spp.). No estimates of abundance, biomass, DC, P/B nor Q/B exist for these species in central Chile. Therefore, we use values for these parameters informed for similar species in comparable ecosystems. B. Fishes Fishes play a very important role in ecosystems as prey, predators and target species for many fisheries. The trophic role of fishes is largely a function of their size, with big fishes normally eating small fishes. However, the high diversity in anatomical and physiological adaptations exhibited by this group leads to a wide range of dietary specializations (i.e., herbivores and carnivores). This results in fishes occupying almost all trophic levels in the food web. We split the fish community in two major categories, bony fishes and chondrichthyans. Bony fishes were split in three main categories based on their habitat and feeding traits: pelagic fish, mesopelagic fish and demersal fish. Due to lack of data we only consider demersal chondrichthyans. Below are the main characteristics of selected fish functional groups. Pelagic fish We split this broad category in three main groups: small- (mainly planktivores), medium- (planktivores and piscivores), and large-sized (mainly piscivores) fish. The constituent species in each category are described below.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  75  Small pelagic fish: common sardine and anchovy We include the two dominant species (in terms of biomass and landings) of small pelagic fish in the study area, common sardine and anchovy. Common sardine is endemic to central Chile, while anchovy presents a wider distribution ranging from southern Ecuador to southern Chile. Both species aggregate in mixed schools distributed in shallow coastal waters (between 0 and 70 meters and <20 nautical miles from the shore). Medium pelagic fish: horse mackerel and hoki We include two species in this sub-category: horse mackerel and hoki (Macruronus magellanicus Lönnberg, 1907). Horse mackerel is distributed in the Pacific Ocean from the eastern boundary off Peru and Chile up to the western boundary off New Zealand. This migratory circuit ranges from the SubTropical Convergence (30ºS) up to the Sub-Antarctic Convergence (48º–50ºS). In central Chile, horse mackerel is the main off-shore fishery resource, being targeted by an important industrial fleet of purse seiners. Hoki has a more restricted geographical distribution than horse mackerel: from central Chile (33ºS) up to austral Chile, reaching the Atlantic off southern Argentina (37ºS; Arana, 1970). Although hoki also reaches deeper waters than horse mackerel (200 to 300 m; Aguayo and Gili, 1984), in central Chile these two species are targeted by the same fleet. Large pelagic fish: swordfish This category is represented by the swordfish (Xiphias gladius). This is a highly migratory and cosmopolitan species distributed in all oceans in the globe, with preference for waters with temperature higher than 13ºC (Barbieri et al., 1998). Swordfish is a carnivorous species that prey mainly on fish and cephalopod species. This allows us to characterize this species as a top predator in the system. However, other life history parameters of swordfish are scarce off central Chile. Therefore, we used mainly input parameters informed for the same or similar species in other ecosystems. Mesopelagic fish Although no direct estimates of the abundance of this group have been conducted off central Chile, mesopelagic fishes are thought to be very abundant, with myctophiids representing the bulk of the biomass of mesopelagic fish (Siefeld et al., 1995). Mesopelagic fishes are important consumers of zooplankton (copepods and euphausiids) and important prey of a wide diversity of predators, especially fish such as horse mackerel and hoki. Therefore, mesopelagic fishes represent an important trophic link between meso- and macrozooplankton and predators located in higher trophic levels (Acuña, 1986). Demersal fish Hake Hake is a relatively large demersal fish that can reach over 80 cm in length. This species is distributed from 23ºS to 47ºS on the shallow continental shelf to the upper continental slope (Aguayo, 1995). In central Chile hake is an opportunistic predator that migrates vertically to mid-waters at night. Hake diet is based on galatheid crustaceans (such as squat lobster), small pelagics (such as anchovy and common sardine), euphausiids and cannibalism (Arancibia et al., 1998; Cubillos et al., 2003). Small squat lobsters constitute most of the diet of young hake (<36 cm LT), while small pelagics are more important in the diet of adult hakes (>36 cm LT) (Arancibia et al., 1998). The fishery targeting Chilean hake is one of the most important fisheries in Chile. The fleet (both industrial and small scale) operates over the continental shelf between 31ºS and 41ºS. Big-eye flounder Big-eye flounder (Hippoglossina macrops Steindachner, 1876) is a demersal fish with a geographic distribution that ranges from central to southern Chile. It is mainly associated with sandy and muddy bottoms in depths of 50 to 120 m (Arancibia, 1989). Big-eye flounder co-occurs in the demersal fish assemblages dominated by Chilean hake, and it is normally found in the by-catch of the fishery targeting hake (Arancibia 1989; 1992). Black conger-eel Black conger-eel (Genypterus maculatus Shneider, 1848) is a demersal species in the industrial and small-  76  Upwelling Ecosystem off Central Chile, Neira and Arancibia  scale trawl fisheries (hake, red squat lobster and pink shrimp). This species inhabits coastal waters from northern Chile (18º25'S) up to southern Chile (47º75'S), in depths between 20 and 150 m. Black congereel is also a co-occurring species in the demersal fish assemblages dominated by hake, and it is normally found in the by-catch of the fishery targeting hake (Arancibia 1989; 1992). Cardinal fish Off central Chile, Cardinal fish (Epigonus crassicaudus) is distributed in depths between 100 and 500 m over the continental shelf and break. This species co-occurs in the demersal fish assemblages dominated by Chilean hake, being normally found as by-catch during summer in the fishery targeting hake (Arancibia 1989; 1992). Pacific sand perch Pacific sand perch (Prolatilus jugularis) is normally associated with rocky and sandy substrata, between 5 and 250 m. Pacific sand perch is also a co-occurring species in the demersal fish assemblages dominated by Chilean hake, and it is normally found in the by-catch of the fishery targeting hake (Arancibia 1989; 1992). Rattail fishes Rattail fishes (Coelorhyncus spp.) inhabit depths between 70 m and 400 m. This group is distributed from northern to austral Chile. Rattail fishes co-occur in the demersal fish assemblages dominated by Chilean hake, and are normally found in the by-catch of the fishery targeting hake (Arancibia 1989; 1992) Small chondrichthyans This group is represented mostly by skates of the Family Rajiidae. In general, chondrichthyans are important due to conservation objectives. Skates are distributed from northern to southern Chile, with Dipterus chilensis being a representative species of the group. Skates are incidentally caught as a by-catch species by the industrial trawlers targeting hake. C. Mollusks In this category we identify Cephalopoda as the main group to be included in the model, since other families such as Bivalvia and Gastropoda are more restricted to sandy and rocky shores. Among cephalopods, squids are largely the most important compared to Octopus spp., which also inhabit rocky shores. We consider small squids and large squids as independent functional groups since they play different roles in the system. Small squids (Loliginidae) are important as prey of several predators (including top predators), while large squids (Omastrephidae) are important predators of fish species (including target species). Small squids Small squids are represented by common squid (Loligo gahi Orbigny), which is an important prey for several predators in the system. Because trophic/population parameters for this groups are not available locally, we use parameters estimated for similar groups in comparable ecosystems. Large squids Large squids are represented by jumbo squid. This is a fast-growing nictimeral species that is distributed along the coast of the Eastern Pacific Ocean (Rodhouse and Nigmatulin, 1996). Trophic studies indicate that this is a voracious and opportunistic predator with a diet based mostly on plankton crustaceans, mollusks (including other cephalopods) and fish. Mesopelagic fish species are the most important prey in jumbo squid diet, with pelagic species such as sardines (Sardinops sp.), anchovies (Engraulis sp.) and mackerel (Scomber japonicus) occupying a secondary position (Rodhouse and Nigmatulin, 1996; Nigmatulin, et al., 2001; Markaida and Sosa-Nishizaki, 2003). Moderate to high cannibalism is common in jumbo squid, but it can account up to 70% of the diet in conditions of scarcity of other prey items (Arancibia et al., 2007). D. Benthic crustaceans Benthic crustaceans are important in the study area (especially in the demersal realm) since they are normal prey of several demersal fish species (Arancibia, 1989). In addition, some species support important fisheries.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  77  Red squat lobster The red squat lobster is an important fishing resource in the study area. However, its biomass decreased significantly in the last decade, down by about 90% with respect to the early 1990s (Canales and Espejo, 2002). Consequently, the fishery has been closed since 2000, without evidence of recovery. Yellow squat lobster Yellow squat lobster is a secondary fishing resource in the study area. Its biomass is relatively low. Because yellow squat lobster is by-catch in the fishery targeting red squat lobster, its fishery is also closed since 2002. Pink shrimp Pink shrimp is also a secondary fishing resource in the study area. Its biomass is relatively low, but it supports a small-scale fishery in central Chile. E. Anelida – Polychaeta Available information indicates that the metazoan component of the benthos off central Chile (Concepción Bay) is dominated, both in number and weight, by Polychaeta. Other groups such as Molluska (Bivalvia and Gastropoda), Crustacea (Amphipoda and Brachyura), and Anthozoa seem to present low abundance in the system (Gallardo, 1979). Polychaetes The polychaete Parapionospio pinnata (Ehlers, 1901) is the dominant species in this functional group. P. pinnata presents euribathic distribution, typically in muddy, lime or sandy substrata (Gallardo, 1979; Carrasco and Arcos, 1980). F. Plankton Copepods The zooplankton community off central Chile is dominated by copepods (in number), which reach higher abundance in coastal areas (<20 nautical miles). The copepod assemblage in the study area is composed of 42 taxa, with the following species dominating: Calanoides patagoniensis, Calanus chilensis, Paracalanus parvus, Acartia tonsa, Centropages brachiatus, Oithona sp. and Oncaea sp. (Núñez, 1995). Euphausiids Euphausiids constitute the main fraction (in weight) of the macro-zooplankton biomass in the study area. They are also a key prey for pelagic (i.e. horse mackerel) and demersal (i.e. hake) fish species. Euphausia mucronata is an endemic krill species in the Humboldt Current system. Phytoplankton The phytoplankton box includes all photosynthetic microorganisms, mainly micro-algae, inhabiting the study area. The phytoplankton box represents the primary producers or first level of the traditional food chain. Off central Chile most of the information about phytoplankton biomass (Chlorophyll-a concentration) show high values (>2 ug L-1; 50 mg m-2) out to 50 km from the coast (Montecino et al., 1998). The primary production off central Chile has been estimated as 19.9 g C m-2 d-1 (Daneri et al., 2000). G. Non-living group Detritus Every Ecopath model requires at least one detritus group. We consider only one detritus box in our model  78  Upwelling Ecosystem off Central Chile, Neira and Arancibia  representing both particulate and dissolved organic matter. This box is the sink for residual mortality for all other functional groups in the model. Since the detritus box corresponds to a ‘dead’ group, no input parameters are entered.  Age structure The model includes age structure in groups where enough data is available to split the group into juvenile and adult stages. According to Arancibia (1987), in Chilean hake, the juvenile group corresponds to age 0 to 3 years old, while the adult group corresponds to ages 4+ years old. In common sardine, anchovy and red squat lobster, we considered that the juvenile and adult groups are individuals of the age group 0–1 year old (recruits) and 1+ years old, respectively.  The Ecopath model The Ecopath model is fully described in many publications (Christensen and Pauly, 1992; Christensen and Walters, 2004; Christensen et al., 2005), and due to space limitations we do not describe the full model in detail. We just present the two master equations in order to help the reader to understand the input parameters needed to run the model. In this application, we used the EwE software version 5.1. The mathematical structure of EwE is based in two master equations: The first master equation describes the mass balance for each group in the model: Q=P+R+U  Eq. 1  where Q is prey consumption, both inside and outside the system (imports); P is production (which must be consumed by predators, exported from the system or contributed to detritus); R is respiration; and U is unassimilated food by predators. The second master equation describes the fate of the production of each group (i) in the model: production = catches + predation mortality + biomass accumulation + net migration + other mortality or, more formally, Pi = Yi + Bi * M2 + Ei + BAi + Pi * (1-EE)  Eq. 2  where i is a model group; Pi is the total production rate of (i); Yi is total fishery catch rate of (i); M2i is the total predation rate for the group (i); Bi the biomass of the group; Ei the net migration rate (emigration immigration); BAi is the biomass accumulation rate for (i); M0i = Pi (1-EEi) is the other mortality rate for (i); and EEi is the ecotrophic efficiency of (i), which represents the total fraction of the production of i that is either eaten by predators or exported from the system. These lead to the following linear equation: Bi * P/Bi + EEi - ∑j (Bj * Q/Bj * DCij) - EXi = 0  Eq. 3  where j indicates any of the predators of (i); P/Bi is the production of (i) per biomass unit (equivalent to total mortality Z under steady-state conditions, sensu Allen, 1971); Q/Bi is the consumption by (i) per biomass unit; DCij is the fraction of (i) in the diet of (j) (in mass units); and EXi are the exports of i (by emigration or yields). This structure defines the input parameters needed to complete the model. Each group requires estimates of B, P/B and Q/B ratios, DCij, EXi, assimilation and EEi. Nevertheless, one of the parameters (B, P/B, Q/B or EE) can remain unknown for each group, since it can be estimated (together with respiration) from the solutions of the system of linear equations. Values of the gross efficiency of food conversion (GE), which corresponds to the production/consumption ratio (P/Q), can be used as alternative inputs to Q/B. For the phytoplankton group it is not necessary to enter Q/B or P/Q values, since this is an autotrophic group. Data sources and estimation methods used to estimate input parameters are presented in Table 1. In absence of further information, we assumed steady state conditions for each group (i) in 2000, i.e. BAi=0 and Ei=0.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  79  Annual Bi in big-eye flounder, black conger-eel, Pacific sand, perch cardinal fish, rattail fish and skates, which are by-catch species in the hake fishery, was estimated as follows:   C Bi = Bhake ∗  i  C hake      Eq. 4  where Bhake is the biomass of hake (Payá et al., 2001); Ci is the yield of the species i during a research cruise carried out to estimate the hake biomass in 2000; Chake is the yield of hake during the same cruise. We assumed that all species had the same response to the Chilean hake trawling fishing gear. Annual Yi for the same groups, were estimated as follows:   C  Yi = Yhake ∗  i   Chake   Eq. 5  where Yhake is the annual landing of hake obtained from the National Fishery Service (SERNAPesca, 2001).  Main assumptions when running the model When running the EwE model, B, P/B, Q/B or EE can be unknown for any group i, since if the other three parameters are entered, EwE will estimate the unknown parameter when solving the system of equations that define the model. We use this model capability to estimate Bi in those groups where no data was available (see Table 1). In small squids, mesopelagic fish, red squat lobster (j), euphausiids and copepods, B was estimated by EwE under the assumption that EE=0.999 for all these groups. This value of EE implies that EwE calculates the minimum biomass required to sustain predators of these groups (including the fishery). The biomass of large squids was calculated assuming EE=0.5. Following Jarre-Teichmann et al. (1998), B in gelatinous zooplankton was estimated assuming that EE=0.15. In phytoplankton, B was estimated assuming EE=0.8 in order to restrict the values of P obtained by EwE in the range of values informed for the marine system off central Chile (Daneri et al., 2000). Considering that for central Chile no B values have been estimated for cetaceans or marine birds and that EwE cannot estimate biomass in top predators, we assumed that the B of cetaceans and marine birds in the upwelling system off central Chile are similar to those calculated in other upwelling systems (Jarre et al., 1989; Jarre-Teichmann et al., 1998).  Balancing the model The model was balanced by checking the values of EEi and GEi. Obviously, EEi must be between 0 and 1, while GEi must lie between 0.1 and 0.35 (Christensen and Pauly, 1992). When inconsistent values of either EEi or GEi were found, then we perform changes in the input parameters (i.e., Bi, Pi/Bi o DCij) following criteria presented by Christensen et al. (2005) until we obtained acceptable outputs, i.e. EE<1 and 0.1<GE<0.35 for each group i. Table 2 presents input and output parameters of the balanced Ecopath model representing the upwelling system of central Chile, year 2000. Table 3 presents the diet composition of predators in the same model.  80  Upwelling System off Central Chile, Neira and Arancibia  Table 1. Functional groups considered in the modelling of the marine system off central Chile, year 2000 and the data used to parameterize the model. Key: (j)=juveniles; (a)=adults; (EC1)= equation 1, see text; (EC2) equation 2, see text; (SA)=Stock assessment; (OR)=Official Report from National Fisheries Service Statistic Yearbooks; (SCA)=Stomach content analysis; (GKS)=General knowledge of the same species/gender/group. Group/Parameter Bi Pi/Bi Qi/Bi Yi DC EEi t·km-2 t·km-2·year-1 year-1 year-1 1. cetaceans 20 and 21 21; 26 7; 26 24; 26 2. sea lions 15 21 20 CGE based on 15 3. marine birds 20 and 21 20 20 20; 21 4. hake (j) SA; 27 2 2 27 SCA; 3 5. hake (a) SA; 22 and 27 2 2 OR; 29 SCA; 3 OR; 29 SCA; 4 6. common sardine (j) SA; 28 Cubillos, pers. com* OR; 29 SCA; 4 7. common sardine (a) SA; 28 Cubillos, pers. com* OR; 29 SCA; 4 8. anchovy (j) SA; 28 Cubillos, pers. com* OR; 29 SCA; 4 9. anchovy (a) SA; 28 Cubillos, pers. com* 10. small squids 21 OR; 29 23; 31 0.999 11. large squids 21 13 OR; 29 13 0.500 12. mesopelagic fish 18 21 1 0.999 13. red squat lobster (j) GKS based on 31 31 GKS based on 31 0.999 14. red squat lobster (a) SA; 10 31 31 OR; 29 GKS based on 31 15. yellow squat lobster SA; 17 31 31 OR; 29 GKS based on 31 16. pink shrimp SA; 9 31 31 OR; 29 GKS based on 31 20 Cubillos, pers. com*. SCA; 3 17. horse mackerel SA; 11 Cubillos, pers. com* 18. hoki SA; 28 21 OR; 29 SCA; 12 19. swordfish 21 OR; 29 SCA; 6 20. black conger-eel Eq. 5 3; 25 3; 25 Eq. 6 3; 25 21. rattail fish Eq. 5 3; 25 3; 25 Eq. 6 3; 25 22. big-eye flounder Eq. 5 3; 25 3; 25 Eq. 6 3; 25 3; 25 3; 25 Eq. 6 3; 25 23. cardinal fish Eq. 5 24. Pacific sand perch Eq. 5 3; 25 3; 25 Eq. 6 3; 25 25. small chondrichthyans Eq. 5 3; 25 3; 25 Eq. 6 3; 25 26. polychaetes 8 8 5 GKS based on 8 27. gelatinous zooplankton 21 21 21 0.150 28. copepods 16 19 0.999 29. euphausiids 19 19 0.999 30. phytoplankton 14 0.800 31. detritus -  GEi  18 18 18 18 30  21  19 19  1=Amstrong et al. (1991); 2=Arancibia et al. (1998); 3=Arancibia et al., (2002); 4=Arrizaga et al. (1993); 5=Arreguin-Sánchez et al. (1993); 6=Barbieri et al. (1998); 7=Browder (1993); 8=Carrasco and Arcos (1980); 9=Canales (2002); 10=Canales and Espejo (2002); 11=Córdova et al. (2000); 12=Cubillos et al. (1998); 13=Cubillos et al. (2004); 14=Daneri et al., 2000; 15=Doppler Ltda. (1997); 16=Escribano and McLaren (1999); 17=Espejo and Canales (2002); 18=Hewitson and Crushak (1993); 19=Hutchings et al. (1991); 20=Jarre et al. (1989); 21=Jarre-Teichmann et al. (1998); 22=Lillo et al. (2001); 23=Lipinski (1992); 24=Majluf and Reyes (1989); 25=Neira et al. (2004); 26=Pauly et al. (1998b); 27=Payá et al. (2001); 28=Quiñones et al. (2002); 29=SERNAPesca (2001); 30=Shannon and Jarre-Teichmann (1999); 31=Wolf (1994). * Luis Cubillos (Universidad de Concepción, P.O. Box 160-C, Concepción, Chile; Email: lucubillos@udec.cl)  -  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  Table 2. Input and output (bold) parameters of the balanced Ecopath model representing the upwelling system off central Chile, year 2000. TL=trophic level; B=biomass; P/B=production:biomass ratio; Q/B=consumption:biomass ratio; Y=landings; EE=ecotrophic efficiency; GE=gross efficiency; F=fishing mortality; M2=predation mortality; M0=other mortalities; j=juveniles; a=adults. Group/Parameter Q/B Y M2 M0 TL B P/B EE GE F ton·km-2·year-1 ton·km-2 year-1 year-1 year-1 year-1 year-1 1. cetaceans 4.41 0.0070 0.600 10.000 0.000 0.167 0.060 0.000 0.100 0.500 2. sea lion 3.93 0.0720 0.250 20.000 0.000 0.381 0.013 0.000 0.095 0.155 3. marine birds 3.58 0.0650 0.500 20.000 0.000 0.000 0.025 0.000 0.000 0.500 4. hake (j) 3.35 7.755 2.500 8.323 0.064 0.990 0.300 0.008 2.467 0.025 5. hake (a) 3.92 12.189 0.456 5.159 2.143 0.668 0.088 0.176 0.129 0.151 6. common sardine (j) 2.03 41.355 1.453 14.530 8.510 0.309 0.100 0.206 0.243 1.005 7. common sardine (a) 2.03 14.600 1.875 18.750 4.594 0.275 0.100 0.315 0.201 1.359 8. anchovy (j) 2.03 23.971 0.703 7.030 3.650 0.613 0.100 0.152 0.279 0.272 9. anchovy (a) 2.03 14.631 2.120 21.200 4.520 0.240 0.100 0.309 0.201 1.611 10. small squids 3.73 3.362 3.500 10.606 0.000 0.999 0.330 0.000 3.497 0.003 11. large squids 4.48 6.413 1.750 5.303 0.000 0.500 0.330 0.000 0.875 0.875 12. mesopelagic fish 3.40 50.299 1.200 12.000 0.000 0.999 0.100 0.000 1.199 0.001 13. red squat lobster (j) 2.00 0.230 5.900 18.000 0.000 0.999 0.328 0.000 5.894 0.006 2.00 0.558 3.569 12.500 0.080 0.999 0.286 0.143 3.422 0.004 14. red squat lobster (a) 15. yellow squat lobster 2.00 0.077 3.569 11.600 0.059 0.782 0.308 0.766 2.023 0.780 16. pink shrimp 2.00 0.400 2.500 12.000 0.089 0.466 0.208 0.223 0.942 1.335 17. horse mackerel 3.52 23.980 0.564 14.200 3.163 0.359 0.040 0.132 0.070 0.362 18. hoki 3.77 21.900 0.528 5.280 1.509 0.705 0.100 0.069 0.303 0.156 19. sword fish 4.66 0.640 0.500 5.000 0.240 0.750 0.100 0.375 0.000 0.125 20. black conger eel 3.53 0.300 0.700 3.500 0.068 0.351 0.200 0.227 0.019 0.454 21. rattail fish 3.00 2.282 0.700 3.500 0.001 0.999 0.200 0.000 0.699 0.001 22. big-eye flounder 3.00 0.200 0.700 3.500 0.002 0.014 0.200 0.010 0.000 0.690 23. cardinal fish 3.50 6.830 0.700 3.500 0.115 0.999 0.200 0.017 0.682 0.001 24. Pacific sand perch 3.57 0.045 0.700 3.500 0.003 0.095 0.200 0.067 0.000 0.633 25. skates 3.00 0.253 0.362 2.413 0.012 0.131 0.150 0.047 0.000 0.315 2.410 15.900 0.000 0.000 0.152 0.000 0.000 2.410 26. polychaetes 2.00 1.886 27. gelatinous zooplankton 2.63 7.774 0.584 1.420 0.000 0.150 0.411 0.000 0.088 0.496 28. copepods 2.25 80.383 45.000 154.519 0.000 0.999 0.291 0.000 44.955 0.045 29. euphausiids 2.50 66.159 13.000 31.707 0.000 0.999 0.410 0.000 12.987 0.013 30. phytoplankton 1.00 347.971 120.000 0.000 0.300 0.000 36.000 84.000 31. detritus 1.00 1000 0.000 0.001 -  81  82  Upwelling System off Central Chile, Neira and Arancibia  Table 3. Diet composition of predators in the balanced Ecopath model representing the upwelling system off central Chile, year 2000.  0.100 0.050 0.284 0.100 0.165 0.101 0.040 0.040 0.120  0.118 0.042 0.070 0.042  0.015 0.008 0.1  0.050  0.5  0.2  0.4  0.5  0.8  0.6  1.0  1.0  1.0  29. euphausiids  26. polychaetes  25. skates  24. Pacific sand perch  23. cardinal fish  22. big-eye flounder  21. rattail fish  20. black conger eel  19. sword fish  0.013 0.036  28. copepods  0.250 0.200 0.180 0.065 0.110 0.062  27. gel. zooplankton  0.06 0.07 0.1 0.03 0.06 0.04 0.15 0.15 0.1  18. hoki  16. pink shrimp  0.01 0.1  17. horse mackerel  15. yellow squat lobster  14. red squat lobster (a)  0.088 0.029  13. red squat lobster (j)  11. large squids  0.050  12. mesopelagic fish  10. small squids  9. anchovy (a)  8. anchovy (j)  0.170  7. common sardine (a)  5. hake (a)  0.040  6. common sardine (j)  4. hake (j)  3. marine birds  2. sea lion  Prey 1 cetaceans 2 sea lion 3 marine birds 4 hake (j) 5 hake (a) 6 com sardine (j) 7 com sardine (a) 8 anchovy (j) 9 anchovy (a) 10 small squids 11 large squids 12 mesopel. fish 13 red s. lobst. (j) 14 red s lobst. (a) 15 yellow s. lobst. 16 pink shrimp 17 h. mackerel 18 hoki 19 sword fish 20 bl. conger eel 21 rattail fish 22 big-eye flound. 23 cardinal fish 24 P. sand perch 25 skates 26 polychaetes 27 gel. zooplankt. 28 copepods 29 euphausiids 30 phytoplankt. 31 detritus Imports Sum  1. cetaceans  Predator  0.065  0.050 0.002 0.002 0.300 0.002 0.001 0.002 0.006  0.210 0.100  0.198 0.206  0.025  0.035 0.035 0.180  0.010 0.010 0.004 0.107 0.002  0.100  0.030 0.110  0.020 0.300 0.020  0.002 0.210 0.023  0.480 0.500  0.124 0.004 0.002  0.036  0.030  0.113  0.500 0.04 1.00 1.000  1.000  0.165 1.000  0.505 0.008 1.000  0.02  0.02  0.02  0.02  0.98  0.98  0.98  0.98  1.00  1.00  1.00  1.00  0.44 0.10  1.00  0.002 0.002 0.971  0.4 0.6 0.206 1.000  1.0  1  1  1  1  1  1  1  1 1.000  0.750  0.5  0.679 1  1.000  1.000  0.838 1.000  0.860 1.000  0.660 1.000  0.5 1.0  0.256 1.000  0.765 1.000  1  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  83  REFERENCES Acuña, E. 1986. El recurso mictófidos (Pisces, Myctophidae): antecedentes en aguas chilenas y marco de referencia para su investigación. In: La Pesca en Chile. P. Arana (Ed.) Valparaíso, Chile, pp. 315-339. Aguayo-Lobo, A., Torres Navarro, D. and Acevedo Ramírez, J. 1998. Los mamíferos de Chile: I. Cetácea. Serie Científica INACH, 48: 19159. Aguayo, M. and Gili, R. 1984. Estudio de la edad y el crecimiento de la merluza de cola (Macruronus magellanicus, Lönnberg) (Gadiformes–Gadidae). Investigaciones Pesqueras (Chile), 31: 47–57. Aguayo, M. 1995. Biology and fisheries of Chilean hakes (M. gayi and M. australis). In: Hakes, biology fisheries and markets. J. Alheit, T. Pitcher (eds). Chapman and Hall, UK, pp. 305-338. Allen, K. R. 1971. Relation between production and biomass. Journal of Fisheries Research Board of Canada, 28: 1573-1581. Amstrong, M.J., James, A.G. and Valdés Szeinfeld, S. 1991. Estimates of annual consumption of food by anchovy and other pelagic fish species off South Africa during the period 1984-1998. South African Journal of marine Science, 11: 251-266. Arana, P. 1970. Nota sobre la presencia de ejemplares de merluza de cola (Macruronus magellanicus, Lönnberg) frente a la costa de Valparaíso. Investigaciones Marinas (Chile), 1(3): 55-68. Arancibia, H. 1992. Distribution patterns of the demersal fish assemblage off Central Chile. Biología Pesquera (Chile), 21: 43-53. Arancibia, H. 1989. Distribution patterns, trophic relationships and stock interactions in the demersal fish assemblage off Central Chile. Dr. Rer. Nat. Thesis, University of Bremen, RFA, 221 pp. Arancibia, H. 1987. On the application of multivariate analysis in the determination of "ontogenetic trophic units" in the Chilean hake, Merluccius gayi (Guichenot, 1848). ICES Demersal Fish Committee, Ref. Statistics Committee, C.M. 1987/G:6. Arancibia, H. and Neira, S. 2003. Simulación de cambios en la biomasa de los principales recursos pesqueros de Chile central (V – IX región) bajo el marco de la ley de pesca N° 19.713 y sus posteriores modificaciones. Informe Final. Universidad de Concepción, 53 pp. Arancibia, H., Barros, M., Neira, S., Markaida, U., Yamashiro, C., Icochea, L., Salinas, C., Cubillos, L., Ibáñez, Ch., León, R., Pedraza, M., Acuña, E., Cortés, A., and Kesternich, V. 2007. Informe Final Proyecto FIP 2005-38. Análisis del impacto de la jibia en las pesquerías de peces demersales. Universidad de Concepción / Universidad Católica del Norte, 299 p. + Anexos. Arancibia, H., Neira, S., Christensen, V., Olson, R., Cubillos, L., Quiñones, R., Gatica, C., and Medina, M. 2002. Enfoque metodológico para el análisis ecosistémico en la administración de pesquerías de la zona central de Chile. Pre-informe Final Proyecto FIP-2001-29. Universidad de Concepción, 266 pp. Arancibia, H., Catrilao, M. and Farías, B. 1998. Evaluación de la demanda de alimento en merluza común y análisis de su impacto en prereclutas. Informe Final Proyecto FIP 95-17. Universidad de Concepción, 98 pp. Arreguín-Sánchez, F., Seijo, J.C. and Valero-Pacheco, E. 1993. An application of Ecopath to the Continental Shelf ecosystem of Yucatan, Mexico. In: Trophic models of acuatic ecosystems. Christensen, V. and D. Pauly. (Eds.). ICLARM Conference Proceedings, Nº 26, pp. 269-278. Arrizaga, A., Fuentealba, M., Espinoza, C., Chong, J. and Oyarzún, C. 1993. Hábitos tróficos de dos especies de peces pelágicos: Strangomera bentincki (Norman, 1936) y Engraulis ringens Jenyns, 1842 en el litoral de la Región del Bio-Bio, Chile. Boletín de la Sociedad de Biología de Concepción (Chile), 6-4: 27-35. Avaria, S. and Muñoz, P. 1982. Producción actual, biomasas y composición específica del fitoplancton de la bahía de Valparaíso en 1979. Revista de Biología Marina de Valparaíso, 18: 129-157. Barbieri, M.A., Canales, C., Correa, V., Donoso, M., González, A., Leiva, B., Montiel, A., and Yáñez, E. 1998. Development and present State of the Sword fish, Xiphias Gladius, fishery in Chile. pp. 1-10. In: Biology and Fisheries of sword fish, Xiphias gladius. Barret, I., O. Sosa-Nishizaki and N. Bartoo (Eds). Papers from the International Symposium on Pacific sword fish, Ensenada, Mexico 11-14 December 1994. Bax, N. J. 1991. A comparison of the fish biomass flow to fish, fisheries, and marine mammals in six marine ecosystems. ICES Mar. Sci. Symp., 193: 217–224. Bax, N. J. 1998. The significance and prediction of predation in marine fisheries. ICES Journal of Marine Science, 55(6): 997-1030. Botsford, L., Castilla, J.C. and Peterson, C.H. 1997. The management of fisheries and marine ecosystems. Science, 277: 509-515. Browder, J.A. 1993. A pilot model of the Gulf of Mexico Continental Shelf. In: Trophic models of aquatic ecosystems, Christensen, V. and D. Pauly. (Eds.), ICLARM Conference Proceedings, Nº 26, pp. 279-284. Camus, P. 2001. Biogeografía marina de Chile continental. Revista Chilena de Historia Natural, 74: 587-617. Canales, C. 2002. Informe de avance Proyecto “Investigación CTP camarón nailon 2002” Diagnóstico actualizado del estado de situación del recurso camarón nailon 2001. 32 pp. + 3 Anexos. Canales, C. and Espejo, B. 2002. Informe de avance Proyecto “Investigación CTP Langostino colorado 2002” Diagnóstico actualizado del estado de situación del recurso langostino colorado. 27 pp. + 6 Anexos. Carrasco F. and Arcos, D. 1980. Estimación de la producción secundaria de Parapionospio pinnata (Spionidea, Polychaeta) frente a Bahía de Concepción, Chile. Boletín del Instituto Oceanográfico de Sao Paulo, 29 (2): 79-82.  84  Upwelling System off Central Chile, Neira and Arancibia  Christensen, V. and Pauly, D. 1992. ECOPATH II. A software for balancing steady state ecosystem models and calculating network characteristics. Ecological Modelling, 61: 169-185. Christensen, V. and Walters, C.J. 2004. Ecopath with Ecosim: methods, capabilities and limitations. Ecological Modelling, 172: 109-139. Christensen, V. and McLean, J.L. 2004. Placing Fisheries in their ecosystem context. Conference Proceedings. Ecological Modelling, 172 (2-4): 103-440. Christensen, V. and Pauly, D. 1993. Trophic models of aquatic ecosystems. ICLARM Conference Proceedings, Nº 26, 390 pp. Christensen V., Walters, C.J. and Pauly, D. 2005. Ecopath with Ecosim: a User’s guide. Fisheries Centre, University of British Columbia, Vancouver. November 2005 edition, 154 p. Córdova, J., Barbieri, Mª A., Miranda, H., Espejo, M., Rojas, M., Núñez, S., Ortiz, J., Catasti, V., Valenzuela, V., Braun, M., Miranda, L., Cuevas, C., Rebolledo, H. and Galindo, G. 2000. Evaluación hidroacústica del recurso jurel en la zona económica de Chile. PreInforme Final Proyecto FIP Nº 2000-03, 110 pp. Cubillos, L., Hernández, A., Vilugrón, L., Miranda, L., Alarcón, R., Pino, C., Sepúlveda, A. and Vásquez, G. 1998. Estudio Biológico pesquero de merluza de cola en el área de distribución de la pesquería pelágica centro-sur. Informe Final Proyecto FIP Nº 96-19. 158 pp. Cubillos, L., Rebolledo, H. and Hernández, A. 2003. Prey composition and estimation of Q/B for the Chilean hake, Merluccius gayi (Gadiformes-Merluccidae), in the central-south area off Chile (34º-40ºS). Arch. Fish. Mar. Res. 50, 271-286. Cubillos, L.C., Ibáñez, C., González, C. and Sepúlveda, A. 2004. Pesca de Investigación: Pesca de jibia (Dosidicus gigas) con red de cerco entre la V a X Regiones, 2003. Instituto de Investigación Pesquera, 48 pp. Cushing, D.H. 1982. Climate and Fisheries. Academic, London/New York. Daneri, G., Dellarossa, V., Quiñones, R., Jacob, B., Montero, P. and Ulloa, O. 2000. Primary production and community respiration in the Humboldt Current System off Chile and associated oceanic areas. Marine Ecology Progress Series, 197: 41-49. Doppler, Ltda. 1997. Censo poblacional del lobo marino común en el litoral de la V a IX Regiones. Informe final Proyecto FIP-IT/96-51, 218 pp. Elton, C. S. 1927. Animal ecology. Sidgwick & Jackson, London. Escribano, R. and McLaren, I. 1999. Production of Calanus chilensis in the upwelling area of Antofagasta, northern Chile. Marine Ecology Progress Series, 177: 147-156. Escribano, E., Fernández, M. and Aranís, A. 2003. Physical-chemical processes and patterns of biodiversity of the Chilean eastern boundary pelagic and benthic marine ecosystems: an overview. Gayana, 67(2):190-205. Espejo, V. and Canales, C. 2002. Informe de avance Proyecto “Investigación CTP Langostino amarillo 2002” Diagnóstico actualizado del estado de situación del recurso langostino amarillo, 26 pp. + 4 Anexos. FAO 2003. The ecosystem approach to fisheries. FAO Technical Guidelines for Responsible Fisheries. No. 4, Suppl. 2. Rome, FAO. 2003. 112 p. Frank, K. T., Choi, J.S., Petrie, B. and Leggett, W.C. 2005. Trophic cascades in a formerly cod dominated ecosystem. Science, 308: 16211623. Gallardo, V.A. 1979. El bacterioplancton de la plataforma continental de la costa sur-occidental de Sudamérica: un reciente descubrimiento. En. Memorias del Seminario sobre ecología bentónica y sedimentación de la Plataforma Continental del Atlántico Sur. UNESCO, Montevideo: 259-267. Hewitson, J. and Crushak, R.A. 1993. Production and consumption by planktivores fish in the northern Benguela ecosystem in the 1980s. South African Journal of marine Science, 13:1031-1050. Hutchings, L., Pillar, S.C. and Verheye, H. 1991. Estimates of standing stock, production and consumption of meso- and macrozooplankton in the Benguela ecosystem. South African Journal of marine Science, 11: 499-512. Jarre, A, Muck, P. and Pauly, D. 1989. Interactions between fish stocks in the Peruvian upwelling ecosystem. ICES Marine Science Symposium, Paper Nº 27, ICLARM Contribution Nº 563, 23 pp. Jarre-Teichmann, A., Shannon, L., Moloney, C. and Wickens, P. 1998. Comparing trophic flows in the Southern Benguela to those in other upwelling ecosystems. In: Pillar, S., Moloney, C., Payne, A., Shillington (Eds.). Benguela Dynamics. South African Journal of marine Science, 19: 391-414. Lillo, S., Ojeda, V., Olivares, J., Tascheri, R., Braun, M., Núñez, S., Ortiz, J. and Torres, P. 2001. Evaluación acústica de merluza común en la zona centro sur, año 2000. Informe final Proyecto FIP Nº 2000-04. 126 p. +29 tablas, 57 figs.+ 2 anexos. Lindeman, R. 1942. The trophic-dynamic aspect of ecology. Ecology, 23: 399-418. Link, J.S. 2002a. What does ecosystem-based fisheries management mean?. Fisheries, 27 (4): 21. Link, J.S. 2002b. Ecological considerations in fisheries management: What does it matter?. Fisheries, 27 (4): 10-17. Lipinski, M.R. 1992. Cephalopods and the Benguela ecosystem; trophic relationships and impact. In: Benguela trophic functioning, Payne, A.J.L., Brink, K.H., Mann, K.H. and R. Hilborn (Eds.), South African Journal of Marine Science, 12: 791-802. MacPhaden, M. 1999. Genesis and evolution of the 1997-1998 El Niño. Science, 283:950-954. Majluf, P. and Reyes, J. 1989. The marine mammals of Peru: a review. 344-363. En: D. Pauly, P. Muck, J. Mendo and I. Tsukayama (Eds).  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  85  The Peruvian upwelling ecosystem: dynamics and Interactions. ICLARM conference Proceedings 18, 483 pp. Markaida, U. and Sosa-Nishizaki, O. 2003. Food and feeding habits of jumbo squid Dosidicus gigas (Cephalopoda: Ommastrephidae) from the Gulf of California, Mexico. J. Mar. Biol. Assoc. U.K., 83 (2003), pp. 507–522. Meléndez, R. 1984. Alimentación de Merluccius gayi (Guichenot) frente a Chile central (32°05’S-36°50’S). Bol. Mus. Nac. Hist. Nat. Chile, 54 p. Miranda, L., Hernández, A., Sepúlveda, A. and Landaeta, M. 1998. Alimentación de jurel y análisis de la selectividad en la zona centro-sur de Chile. In: Arcos, D. (Ed.), Biología y Ecología del Jurel en Aguas Chilenas. Instituto de Investigación Pesquera, Talcahuano-Chile, pp. 173– 187. Montecino, V., Paysant, L., Barría, I., and Sepúlveda, A. 1998. Space scales of variability of chlorophyll-a off central Chile (between 33ºS40ºS): ground truth of April-May 1984 coastal zone color scanner (CZCS) images. Ocean Optics XIV Kaiua Kona, Hawaii, USA: Extended abstract. Montecinos, A., Purca, S. and Pizarrro, O. 2003. Interannual-to-interdecadal sea surface temperature variability along the western coast of South America. Gophisical Research Letters, 30(11): 1570-1573. Neira, S. and Arancibia, H. 2004. Trophic interactions and community structure in the Central Chile marine ecosystem (33ºS-39ºS). Journal of Experimental Marine Biology and Ecology, 312: 349-366. Neira, S., Arancibia, H. and Cubillos, L. 2004. Comparative analysis of trophic structure of commercial fishery species off Central Chile in 1992 and 1998. Ecological Modelling, 172 (2-4): 233-248. Nigmatulin, Ch.M., Nesis, K.N. and Arkhipkin, A.I. 2001. A review of the biology of jumbo squid Dosidicus gigas (Cephalopoda: Omastrephidae). Fisheries Research, 54: 9-19. Núñez, S. 1995. Variaciones temporales y espaciales sen la asociación de copépodos del sistema de surgencia de Talcahuano (Chile). Tesis de Magister, Universidad de Concepción. 162 pp. + 10 anexos. Osgood, W.H. 1943. The Mammals of Chile. Field Museum of Natural History, Chicago. Zoological Series Vol. 30. 268 pp. Palma, S. and Rosales, S. 1995. Composición, distribución y abundancia estacional del macroplancton de la bahía de Valparaíso. Investigaciones Marinas 23: 49-66. Pauly, D. and Christensen, V. 1995. Primary production required to sustain global fisheries. Nature, 374: 255-257. Pauly, D., Christensen, V., Dalsgaard, J.P.T., Froese, R. and Torres, F. 1998a. Fishing down the marine food webs. Science, 279: 860-863. Pauly, D., Trites, A.W., Capuli, E. and Christensen, V. 1998b. Diet composition and trophic levels of marine mammals. ICES Journal of Marine Science, 55:467-481. Pauly, D., Christensen, V. and Walters, C. 2000. Ecopath, Ecosim, and Ecospace as tools for evaluating ecosystem impacts of fisheries. ICES Journal of Marine Science, 57: 697-706. Pauly, D., Alder, J., Bennett, E., Christensen, V., Tyedmers, P. and Watson, R. 2003. The future of world fisheries. Science 302, 1359-1361. Payá, I., Tascheri, R., Montenegro, C. and Bravo, C. 2001. Investigación CTP Merluza común 2002. Primer Informe de Avance. IFOP, 43 pp. + figuras y anexos. Pikitch, E.K., Santora, C., Babcock, E.A., Bakun, A., Bonfil, R., Conover, D.O., Dayton, P., Doukakis, P., Fluharty, D., Heneman, B., Houde, E.D., Link, J., Livingston, P.A., Mangel, M., McAllister, M.K., Pope, J. and Sainsbury, K.J. 2004. Ecosystem-based fisheries management. Science, 305: 346-347. Quinoñes, R., Aliaga, B., Arancibia, H., Navarrete, I., Barriga, O., Henríquez, G., Palma, C., Dresdner, J., Chávez, C. and Salgado, H. 2002. Proyecto análisis biológico, económico y social de las pesquerías de la VIII Región. Informe de avance proyecto FNDR código BIP N° 20183334-0.181 pp. + 8 Anexos. Rodhouse, P.G. and Nigmatulin, Ch.M. 1996. Role as consumers. Philosophical Transactions. Biological Sciences, Vol. 351, No. 1343, The role of cephalopds in the World’s Oceans. Pp. 1003-1022. SERNAPesca. 2001. Anuario Estadístico de Pesca 2000. Departamento de Información y Estadística Pesquera. Ministerio de Economía, Fomento y Reconstrucción, Chile. 283 pp. Shannon, L.J. and Jarre-Teichmann, A. 1999. A model of trophic flows in the Benguela Upwelling System during the 1980s. South African Journal of Marine Science, 21: 349-366. Siefeld, W., Vargas, M. and Fuenzalida, R. 1995. Peces mesopelágicos frente a la costa norte de Chile (18º25'-21º47'S). Investigaciones Marinas Valparaíso (Chile), 23: 83-97. Siefeld, W. 1999. Estado del conocimiento sobre conservación y preservación de Otaria flavescens (Shaw, 1800) y Arctocephalus autralis (Zimmermann, 1783) en las costas de Chile. Estudios en Oceanologia (Chile), 18: 81-96. Sinclair, M., O’Boyle, R., Burke, D.L. and Peacock, G. 1997. Why do some fisheries survive and other collapse?. In: D. A. Hancock, D. C. Smith, A. Grant, and J. P. Beumer (Eds.). Developing and Sustaining World Fisheries Resources: The State of Science and Management. Second World Fisheries Congress, Brisbane 1996. CSIRO publishing, Melbourne, pp. 23-35. Strub, P., Mesías, J.M., Montecino, V., Ruttlant, J. and Salinas, S. 1998. Coastal ocean circulation off western South America. In: The Sea 11. Brink K.H. and A.R. Robinson (eds). Wiley and Sons, Inc., New York, USA, pp. 273-313. Walters, C., Christensen, V. and Pauly, D. 1997. Structuring dynamic models of exploited ecosystems from trophic mass-balance assessments. Reviews in Fish Biology and Fisheries, 7: 139-172.  86  Upwelling System off Central Chile, Neira and Arancibia  Walters, C.J. and Martell, S.J.D. 2004. Fisheries Ecology and Management. Princeton University Press. Princeton, New Jersey, USA. Wolff, M. 1994. A trophic model for Tongoy Bay. A system exposed to suspended scallop culture (Northern Chile). Journal of Experimental Marine Biology and Ecology, 182: 149-168. Worm, B., Barbier, E.B., Beaumont, N., Duffy, J.E., Folke, C., Halpern, B.S., Jackson, J.B.C., Lotze, H.K., Micheli, F., Palumbi, S.R., Sala, E., Selkoe, K.A., Stachowicz, J.J. and Watson, R. 2006. Impacts of biodiversity loss on ocean ecosystem services. Science, 787-790.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  87  SPATIAL RESOURCES AND FISHERY MANAGEMENT FRAMEWORK IN THE EAST CHINA SEA1 He-Qin Chenga,e, Hong Jianga, Hai-Gen Xub, Jun Wub, Hui Dingb, Will Le Quesnec, and Francisco Arreguín-Sánchezd - State Key Laboratory of Estuarine & Coastal Research, East China Normal University, 200062, Shanghai, China; b - Nanjing Institute of Environmental Sciences, State Environmental Protection Administration of China, 210042, Nanjing, China; c - School of Marine Science and Technology, University of Newcastle, Ridley Building, Newcastle upon Tyne NE1 7RU, UK; d - Centro Interdisciplinario de Ciencias Marinas del IPN, Apdo. Postal 592, La Paz, 23000, Baja California Sur, México e - Email: hqch@sklec.ecnu.edu.cn a  ABSTRACT A unique combination of environmental factors including primary productivity, water currents, water masses, temperature, salinity and nutrient level in the East China Sea (ECS) has supported a highly productive fishery and rich biodiversity. Since the 1950s, the government of the People’s Republic of China has introduced a range of spatial management measures including marine protected areas (MPAs), fishery protected areas (FPAs) and large-scale seasonal spatial fishing closures to protect the ECS marine resources. Based on the constructed ECS Ecopath and Ecosim model, this study focuses on the spatial model (Ecospace) description. The main spatial information consisting of environmental factors in the ECS are briefly reviewed, and basic data input to the ECS Ecospace model, such as habitat, location of MPAs, allocation of fleets and main scenarios are introduced.  INTRODUCTION The East China Sea (ECS) is a typical epicontinental sea and part of the western Pacific Ocean bordered by China, South Korea and Japan (Figure 1). It covers an area of 770,000 km² (Zheng et al., 2003), of which 65% has a water depth less than 200 m. The ECS deepens eastward and southward to a maximum depth of 2,300 m in the Okinawa Trough. Large quantities of land-based nutrients and pollutants flow into the ECS along with large fresh water inputs, mainly from the Changjiang (Yangtze) riverine system. The confluences of the alongshore current, the Yellow Sea cold water mass and the Kuroshio Current provide good fishing grounds in the ECS. The superior geography of the ECS has supported a highly productive fishery and rich biodiversity. Economic development and population growth in China and neighboring countries over the last several decades have led to intensified anthropogenic impacts on fishing stocks and biodiversity in the East China Sea. In recent  Figure 1. Study area of the East China Sea  Cite as: Cheng, H.Q., Jiang, H., Xu, H.G., Wu, J., Ding, H., Le Quesne, W. and Arreguín-Sánchez, F. 2007. Spatial resources and fishery management framework in the East China Sea, p. 87–99. In: Le Quesne, W.J.F., Arreguín-Sánchez, F. and Heymans, S.J.J. (eds.) INCOFISH ecosystem models: transiting from Ecopath to Ecospace. Fisheries Centre Research Reports 15(6). Fisheries Centre, University of British Columbia [ISSN 1198-6727].  1  88  East China Sea, Cheng, Jiang, Xu, Wu, Ding, Le Quesne and Arreguín-Sánchez  years, species outbreaks (such as large jellyfish) in the ECS have been thought to be related to ecosystem and fishery resources deterioration. In response, the Chinese government has introduced a range of spatial management measures including large-scale seasonal spatial closures, offshore fishery boxes and coastal biodiversity protected areas. Based on a previously constructed ECS Ecopath and Ecosim model (Jiang et al, 2006), this report describes the development and parameterization of the present of the ECS Ecospace model. The sources and information used to construct the ECS Ecospace model are presented.  SPATIAL CHARACTERISTICS OF THE ECS ECOSYSTEM The ECS is a vast, semi-enclosed marginal sea. The water depth of most of the ECS continental shelf ranges from 60 to 140 m with an average depth of 72 m. The depth contours of the ECS are almost parallel to the coastline. Many harbors and islands are located to the west of the ECS shelf where the water depth is less than 30 m. Primary productivity, water current, water mass, temperature, salinity and nutrient level of the ECS have important effects on the distribution and productivity of the fishery resources, which can be considered environmental resources.  Primary productivity There is seasonal variation and a bimodal character to the monthly distribution of primary productivity in the ECS. The monthly distribution of primary productivity is at its lowest level in the winter and rises rapidly to its peak in spring. It goes down a little in summer, with a slight increase in autumn (Table 1, Li et al., 2003). The spatial distribution shows that the primary productivity decreases from the Figure 2. Distribution of total primary producnear-shore zone toward the open sea (Figure 2). Primary tivity in the ECS for the year 1998 (g·m-2·yr-1). productivity estimated by SeaWiFS is less than 200 g·m-2·yr-1 (wet weight) southeast of the line from Taiwan northeast to Kyushu Island, is about 200 to 250 g·m-2·yr-1 in the middle part of the ECS and is higher than 250 g·m-2·yr-1 in near-shore waters. The peak values are higher than 400 g·m-2·yr-1 in the Changjiang River estuary and Hangzhou Bay, where the important Zhoushan Fishing Ground is located. Table 1. Monthly variation of primary productivity in g·m-2·yr-1 in the ECS in 1998 (Li et al., 2003). Time -2  g·m ·yr  -1  Jan.  Feb.  Mar.  April  May  June  July  Aug.  Sep.  Oct.  Nov.  Dec.  440.1  519.3  538.9  610.5  681.8  696.1  607.6  534.6  516.0  543.3  522.5  464.7  Hydrodynamic regime Water currents Water currents in the East China Sea are dominated by the Coastal Current and the Kuroshio Current. In the western part, the ECS Coastal Current (ECSCC) flows southward in fall and winter and northward in summer (Su, 2001) (Figure 3). To the east of the ECSCC, the broad shelf is dominated by the Taiwan Warm Current (TWC), which starts from the Taiwan Strait and penetrates into the Yellow Sea and/or flows toward the Tsushima Strait (Su, 2001; Liu et al., 2003) (Figure 3). Year-round northward expansion of the TWC has been observed by hydrological investigations (Su, 2001). The TWC is constrained within a water depth of 50–100 m near the west coast. On the eastern boundary, the Kuroshio Current moves northward, separating the ECS from the open ocean (Figure 3), and its branches intrude into the ECS, forming upwelling water due to topographic deflection (Tang et al., 2000; Liu et al., 2000). This circulation pattern greatly influences the dispersal of sediment and fresh water. During fall and winter, the Changjiang plume is constrained by an intensified coastal current flowing in a narrow band southwards along the coast (Liu et al., 2003) and prohibited by  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  89  the TWC from further eastward expansion to the shelf (Su, 2001). In the flood season (i.e., summer), Changjiang-diluted water may extend northeastward to Cheju Island (Liu et al., 2003). Water masses Kuroshio warm water mixes with the low-salinity coastal water when it rushes onto the continental shelf of the ECS. This mixing produces three main water masses in the ECS: Kuroshio water, mixed modified water and coastal water (Liu et al., 1993). The surface and bottom water masses in the ECS during February and August are given in Figure 4. The Kuroshio Current includes two parts, the Kuroshio Surface Water Mass (K) and the Kuroshio Subsurface Water Mass (KC) on the ECS continental shelf. The K is characterized by high temperature (mean annual temperature 18.0–29.0°C) and high salinity (mean annual salinity 34.2–34.9 ppt). The K is distributed mainly in the south of the ECS and covers an area ¼ the size of the whole ECS; thus its fluctuation and change controls the hydrodynamic condition of the ECS directly. The KC with medium temperature (mean annual temperature 14.4– 20.5°C) and high salinity (mean annual salinity 34.4–34.9 ppt) is under the water layer of the K. The KC occurs widely across the ECS continental shelf from May to November. The mixed modified water masses can be divided into 9 parts, according to geographical location and degree modified: E, EC, T, TC, YE, Y, YC, C and M (Figure 4). The main features of the mixed modified water masses are the great changes in their seasonal distribution and medium temperature and salinity between the Kuroshio and coastal water.  Figure 3. Morphology and current patterns in the ECS. ECSCC - ECS Coastal Current; TWC - Taiwan Warm Current; YSCC - Yellow Sea Coastal Current; YSWC - Yellow Sea Warm Current. Solid black lines indicate year-round current; solid gray lines indicate winter currents; dashed black lines indicate summer currents (from Deng et al., 2006).  Diluted water from the Changjiang and Qiantang Rivers is the main source of the ECS coastal water mass (C) that is characterized by low salinity (lower than 31 ppt). Land-air temperature and river runoff have a significant influence on the temperature of the coastal water masses (mean annual temperature 4.0–28.0°C) and on their salinity. The ECS coastal water occurs as a narrow belt along the coast in autumn and winter, but a finger-like projection expands northeast from the Changjiang estuary across the ECS continental shelf from spring to summer. As the main component of the water column over the ECS continental shelf, ECS surface water (E) features significant seasonal fluctuation of temperature and salinity. It is affected by strong vertical mixing, with a range of temperature of 13.0–19.0°C and salinity of 33.75–34.75 ppt, bordered by a strong front with the low temperature and low salinity of the coastal water mass (C) in the southwest part of the ECS in winter. The E occurs in the southeast of the ECS in spring, but in the fall it occurs back on the main parts of the ECS continental shelf with a temperature range of 19.0–25.0°C and salinity larger than 33.75 ppt. The ECS bottom cold water mass (EC) occurs in the southeast of the ECS continental shelf due to the rushing of the KC in summer. Separated from the K, the Taiwan Strait surface water mass (T) is limited by the submarine bathymetry; it shows obvious seasonal changes of temperature and salinity. It is distributed on the narrow part of the southwest ECS, with temperature of 17.9–19.6°C and salinity of 31.0–34.0 ppt in winter, 28.0–29.0°C and 33.1–34.2 ppt in summer. The Taiwan Strait bottom water mass (TC) occurs in a high salinity (>34 ppt) area generated from the KC in summer. The Yellow Sea surface cold water mass (Y), characterized by low temperature and medium salinity, is distributed in the northwest of this study area. It is mixed vertically during winter at low temperatures (<12.0°C) and higher salinity (31.5–33.0 ppt). The Y expands eastward and also, in part, deeper to the ECS bottom, with a temperature of 22.0°C and a salinity of less than 32.0 ppt during summer. Part of the northwest ECS bottom is covered by Yellow Sea bottom cold water mass (YC) with temperature of 7.0– 14.0°C and salinity of 32.5–34.0 ppt in summer, which strongly impacts the ECS fishery.  90  East China Sea, Cheng, Jiang, Xu, Wu, Ding, Le Quesne and Arreguín-Sánchez  Figure 4. Distribution of water masses in bottom (A & B) and surface (C & D) layers in the ECS. K - Kuroshio surface water; KC - Kuroshio subsurface water; E - ECS surface water; EC - ECS bottom cold water; T - Taiwan Strait surface water; TC - Taiwan Strait bottom water; YE - Yellow Sea and ECS mixed water; Y - Yellow Sea surface cold water mass; YC - Yellow Sea bottom cold water mass; C - ECS coastal water mass; M – mixed and transitional water (Yang et al., 2001; Zheng et al., 2003; Hong and Yang, 2005)  The Yellow Sea and ECS mixed water (YE) is variable, distributed in the northeast part of the ECS in winter. It features strong vertical mixing in temperatures of 8.5–15.0°C and salinity of 32.5–34.0 ppt. Mixed and transitional water (M) normally occurs between the mixed modified water or coastal water mass (C).  Sea Surface Temperature The spatial distribution of mean sea surface temperature (SST) in the ECS shows a decrease from southeast to northwest, which indicates the influence of the Kuroshio (Zeng et al., 2006). The seasonal variation of SST is distinctive, as the ECS is located in subtropical and temperate climate zones, but annual variation is small due to the continual influence of the Kuroshio warm water current. The distribution of yearly variation in SST in the ECS shows the same trend as SST (Figure 5). Specifically, the yearly variation in the main part of the Kuroshio is the smallest (within 8°C); in the Taiwan warm current and Tsushima warm current zones it is slightly larger (8–14°C); in the Yellow Sea warm current zone and Taiwan Strait variation is 14–17°C and 11–17°C, respectively; and the maximum seasonal variation in SST in the Changjiang Estuary and along the northeastern coast of the ECS is 20–26.1°C.  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  91  Sediment types Surface sediments of the East China Sea continental shelf fall into 9 categories (Figure 6): fine medium sand, medium fine sand, fine sand, silty fine sand, muddy silt, fine sandy silt, mud-silt-sand, muddy silt and silty mud, which are divided by the 50–60 m depth contour line into two east and west belts. In the western belt, sediments are predominantly fine ooze and silt coming from the Yangtze River, Qiantang River and Oujiang River. Mud and sand coming from the Yellow Sea form sediments in the vicinity of the Yangtze River estuary and in the northeast part of the East China Sea. Sediments in the east belt consist primarily of fine sand originating from the shallow waters. Surface sediments on the continental shelf range from medium fine sands to silty mud. Sediments in the Okinawa Trough are primarily ooze.  Nutrient levels Marine nutrient levels (including inorganic nitrogen, phosphate and silicate) are closely related to primary productivity (Ning et al., 1985). Recently, eutrophication has been increasing in parts of the ECS, and red tides are increasingly frequent. This has resulted in destroyed habitats and damaged fishery resources in the ECS (Zou et al., 1983). The nutrient levels and changes in the ECS are affected by river runoff in coastal geo- and biochemical processes (Gu, 1991).  Figure 5. Yearly variation in SST in the ECS and Yellow Sea in 2005 (Shen et al., 2006)  The inorganic nitrogen (IN) and nutrient level E (E = COD×IN×IP/4500, µmol/dm3) distribution in the middle part of the ECS are given in Figure 7, where COD is chemical oxygen demand and IP is inorganic phosphorus. Values of E in the ECS are between 0.01 and 2.62 with an average of 0.28, which indicates that most of the ECS is still oligotrophic.  Figure 6. Sediment composition of the East China Sea  92  East China Sea, Cheng, Jiang, Xu, Wu, Ding, Le Quesne and Arreguín-Sánchez  Figure 7. The inorganic nitrogen (A) and nutrient level E (B) distribution in the ECS (Zheng et al., 2003)  SPATIAL MODELLING OF FISHERY MANAGEMENT SCENARIOS IN THE ECS The ECS supports a highly productive fishery and is rich in biodiversity. In 2002 landings from the Chinese ECS accounted for over 7% of reported global landings. Economic development and population growth in China over the last four decades have led to intensified anthropogenic impacts on the ECS, including impact on fishing stocks and biodiversity. In response, the government of the People’s Republic of China has introduced a range of spatial management measures including marine protected areas (MPAs), fishery protected areas (FPAs) and large-scale seasonal spatial fishing closures. In order to examine optimal management with sustainable development and spatial fishery management, a spatial ecosystem model (Ecospace) of the ECS was developed to examine the effects of the current management framework. Ecospace is a dynamic and spatial ecosystem modeling tool that is especially effective for fishery and MPA management exploration (Christensen et al., 2005). Spatial information on environmental factors and resources are integrated into Ecospace in order to develop a non-spatial Ecopath and Ecosim model into a spatially explicit ecosystem simulation. This model is developed to simulate the large-scale spatial closures and offshore fishery boxes in the ECS.  Base map The base map is set up as a grid with 48 columns and 38 rows. The length of each cell is 25.8 km. The northwestern corner of the base map is located at 33°N and 117°W.  Habitat definition Five habitats (Figure 8) have been defined based on the distribution of water mass, water depth, bed sediment, water temperature, salinity and nutrient level. These are: H1: Alongshore current (C): The habitat C covers the alongshore currents and is also characterized by low salinity of 28.4 ppt in summer to 30.4 ppt in winter, remarkable inter-annual variation of water temperature from 11.59°C in winter to 24.5°C in summer, with high turbidity and nutrient levels over the sediments, which are composed of mud, silts and fine sands. The alongshore current area on the continental shelf of the ECS provides spawning grounds for many species with high economical value, such as hairtail (Trichiurus lepturus) and large yellow croaker (Larimichtys crocea). There are 17 national nature reserves and 5 special marine protected areas in this  INCOFISH Ecosystem Models, Le Quesne, Arreguín-Sánchez and Heymans  93  habitat. Since 1955, China established a proh