UBC Community and Partners Publications

Modelling the trophic role of marine mammals in tropical areas: data requirements, uncertainty, and validation Morissette, Lyne 2012

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

Item Metadata

Download

Media
FCRR_2009_17-2.pdf
FCRR_2009_17-2.pdf [ 3.41MB ]
FCRR_2009_17-2.pdf
Metadata
JSON: 1.0074740.json
JSON-LD: 1.0074740+ld.json
RDF/XML (Pretty): 1.0074740.xml
RDF/JSON: 1.0074740+rdf.json
Turtle: 1.0074740+rdf-turtle.txt
N-Triples: 1.0074740+rdf-ntriples.txt
Citation
1.0074740.ris

Full Text

  ISSN 1198-6727  Fisheries Centre Research Reports      2009 Volume 17 Number 2    Modelling the trophic role of marine mammals in tropical areas: data requirements, uncertainty, and validation      Fisheries Centre, University of British Columbia, Canada  Modelling the trophic role of marine mammals in tropical areas: data requirements, uncertainty, and validation     by Lyne Morissette, Jenny Lynn Melgo, Kristin Kaschner and Leah Gerber                    Fisheries Centre Research Reports 17(2) 120 pages © published 2009 by  The Fisheries Centre, University of British Columbia  2202 Main Mall Vancouver, B.C., Canada, V6T 1Z4       ISSN 1198-6727  Fisheries Centre Research Reports 17(2) 2009   MODELLING THE TROPHIC ROLE OF MARINE MAMMALS IN TROPICAL AREAS: DATA REQUIREMENTS, UNCERTAINTY, AND VALIDATION   by Lyne Morissette, Jenny Lynn Melgo, Kristin Kaschner and Leah Gerber  CONTENTS   Page DIRECTOR’S FOREWORD ..........................................................................................................................................1 ABSTRACT............................................................................................................................................................... 2 GENERAL INTRODUCTION ....................................................................................................................................... 3 An Ecosystem Approach to the ‘whales eat fish’ Issue.................................................................................. 3 References ................................................................................................................................................... 4 FOOD WEB MODEL AND DATA FOR STUDYING THE INTERACTIONS BETWEEN MARINE MAMMALS AND FISHERIES IN THE NORTHWEST AFRICAN ECOSYSTEM .................................................................................................................. 6 Abstract........................................................................................................................................................ 6 Introduction ................................................................................................................................................ 6 Materials and methods ............................................................................................................................... 7 Study area .................................................................................................................................................... 7 Harvesting on the ecosystem......................................................................................................................8 Whaling ...................................................................................................................................................8 Fishery.....................................................................................................................................................8 Model description ..................................................................................................................................... 13 Groups 1 - 10. Marine mammals.......................................................................................................... 16 11. Seabirds............................................................................................................................................ 27 12. Large pelagics.................................................................................................................................. 27 13. Mesopelagic predators....................................................................................................................28 14. Bathydemersal predators ...............................................................................................................28 15. Sharks ..............................................................................................................................................29 16. Rays..................................................................................................................................................29 17. Coastal tunas ...................................................................................................................................30 18. Coastal demersals ...........................................................................................................................30 19. Clupeids ........................................................................................................................................... 31 20. Other coastal pelagics ....................................................................................................................32 21. Cephalopods ....................................................................................................................................32 22. Crustaceans.....................................................................................................................................33 23. Benthos ...........................................................................................................................................33 24. Benthic producers ..........................................................................................................................34 25. Zooplankton....................................................................................................................................34 26. Phytoplankton ................................................................................................................................34 27. Detritus............................................................................................................................................34 Balancing the model.................................................................................................................................. 35 Time series data ........................................................................................................................................40 Biomass .................................................................................................................................................40 Effort .....................................................................................................................................................43 Uncertainty analyses.................................................................................................................................44 Sensitivity analysis ............................................................................................................................... 45 ‘Ecoranger’ analysis .............................................................................................................................. 45  Fitting the model to time series data................................................................................................... 46 Simulating the removal of great whales in the ecosystem ................................................................. 46 Discussion ................................................................................................................................................. 46 A new dataset built for Northwest Africa............................................................................................ 46 Strengths and weaknesses of these modelling efforts ........................................................................ 47 Acknowledgements ................................................................................................................................... 47 References ................................................................................................................................................. 47 FOOD WEB MODEL AND DATA FOR STUDYING THE INTERACTIONS BETWEEN MARINE MAMMALS AND FISHERIES IN THE CARIBBEAN ECOSYSTEM................................................................................................................................. 53 Abstract ..................................................................................................................................................... 53 Introduction .............................................................................................................................................. 53 Materials and methods ............................................................................................................................. 55 Study area ............................................................................................................................................. 55 Model description ..................................................................................................................................... 56 Resource exploitation of the ecosystem ...............................................................................................61 Balanced ecosystem model for the Caribbean region ............................................................................. 70 Groups 1 - 10. Marine mammals ......................................................................................................... 70 11. Seabirds ........................................................................................................................................... 79 12. Seaturtles.........................................................................................................................................80 13. Large tunas and billfishes ...............................................................................................................81 14. Small tunas.......................................................................................................................................81 15. Dolphinfish...................................................................................................................................... 82 16. Flyingfish......................................................................................................................................... 83 17. Other offshore predators ................................................................................................................ 83 18. Pelagic sharks ................................................................................................................................. 84 19. Coastal and demersal sharks and rays .......................................................................................... 85 20. Scombrids....................................................................................................................................... 85 21. Small and schooling pelagics ......................................................................................................... 86 22. Reef fishes....................................................................................................................................... 87 23. Coastal predators ...........................................................................................................................88 24. Cephalopods ...................................................................................................................................88 25. Crustaceans and benthos ............................................................................................................... 89 26. Zooplankton ................................................................................................................................... 89 27. Benthic producers .......................................................................................................................... 90 28. Phytoplankton ................................................................................................................................ 90 29. Detritus ........................................................................................................................................... 90 Balancing the model ..................................................................................................................................91 Time series data ........................................................................................................................................ 98 Fishing effort ............................................................................................................................................105 Uncertainty analyses................................................................................................................................107 Sensitivity analysis ..............................................................................................................................107 ‘Ecoranger’ analysis............................................................................................................................ 108 Fitting the model to time series data................................................................................................. 108 Discussion ............................................................................................................................................... 109 Strengths and weaknesses of the model.............................................................................................110 Acknowledgements ...................................................................................................................................111 References .................................................................................................................................................111 APPENDICES ........................................................................................................................................................ 117   A Research Report from the Fisheries Centre at UBC 120 pages © Fisheries Centre, University of British Columbia, 2009   FISHERIES CENTRE RESEARCH REPORTS ARE ABSTRACTED IN THE FAO AQUATIC SCIENCES AND FISHERIES ABSTRACTS (ASFA) ISSN 1198-6727  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  1 DIRECTOR’S FOREWORD This document demonstrates the importance of making the best use of local data while addressing critical issues in marine ecology and fisheries management. So far only few attempts have been made to describe the foodwebs in data-scarce areas of the world such as Northwest Africa or the Caribbean. However, whenever such attempts have been made, one realizes that hidden treasures do exist in such cases. Moreover, when data is integrated into a foodweb model, we gain a lot of knowledge by examining that whole ecosystem and determining where more data and research are needed, which can be very fulfilling.  Indeed, once we have a complete description of the whole food web, not only can we learn about the dynamics of the system and the ecology of each species included in it, but we can also discover indirect effects that the complexity of these systems holds. These indirect effects are only seen when examining the whole ecosystem structure, and can lead to counter-intuitive impacts on some trophic groups by others. In the case of marine mammals and their interaction with fish and fisheries, these effects can become very important conservation matters.  Marine mammals are part of ecosystem complexity, and often, they are not considered in models built for fisheries management. By applying an Ecopath with Ecosim approach, this report demonstrates that these animals can be incorporated in marine ecosystem models, and that their role in structuring the ecosystem is important. Thus, assessing their impact on the ecosystem or their interactions with fisheries requires a holistic overview of the ecosystem complexity, not only an assessment of predator-prey issues involving two or three species.  The ‘whales eat fish’ issue is a major point of discussion internationally, and more importantly in tropical countries who are highly depending on marine resources and where fisheries have encountered severe declines over the last decades. When addressing a very socio-politically complex issue such as the interactions between marine mammals and fisheries, it is crucial to make the best use of the data available in tropical countries and to integrate that into an ecosystem approach. A key contribution of this report is that it represents the most updated database on the biology and ecology of all marine organisms in two tropical ecosystems: the Caribbean and Northwest Africa. This was not only collected through an in-depth literature review, but also validated by many local experts in both areas during workshops held in Bridgetown, Barbados, and Dakar, Senegal.  Once again, the current report demonstrates how one of the key areas of research at our Centre can be used to provide insights into fisheries that would otherwise not be studied at the ecosystem level because of the usual excuse of “no data”.  Rashid Sumaila May 12, 2009   Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  2 ABSTRACT This Fisheries Centre Research Report includes two papers that describe whole-ecosystem models of two tropical breeding areas for baleen whales: Northwest Africa and the Caribbean. A mass-balance model, sources of data, and derivations of model parameters are detailed for each region. Exploration of the different sources of uncertainty and their effects on the modelling outcomes are also provided. Analyses based on these models examine the potential competition between whales and fisheries for marine resources.    Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  3 GENERAL INTRODUCTION AN ECOSYSTEM APPROACH TO THE ‘WHALES EAT FISH’ ISSUEa Lyne Morissette Arizona State University, Institut des Sciences de la mer de Rimouski & Fisheries Centre, The University of British Columbia 2202 Main Mall, Vancouver BC V6T 1Z4, Canada Email: lyne.morissette@globetrotter.net  Whales are large animals that are often seen as important predators in the world’s oceans. Recently, some studies have even suggested that whales could be the cause of declined fish populations and that consequently they should be culled (Anonymous 2001a, Komatsu and Misaki 2003; also documented in Struck 2001; Jackson 2007; Holt 2007). For example, Tamura (2003) proposed that the total annual prey consumption by cetaceans in the world was three to six times the amount taken in marine fisheries. The idea of whaling is then propagated as a solution to increase food supply available for fisheries (Anonymous 2001a) This is often proposed that whales are responsible for the worldwide decline of fisheries resources (Komatsu and Misaki 2003) and that an eventual surplus of biomass (the 249-436 million tons of fish consumed by whales) could be directly available for human consumption if the abundance of cetaceans was to be reduced (Anonymus 2001b). Living marine organisms have evolved together in an intricate web of feeding relationships structured on a template of these complex ocean habitats. However, fisheries are now depleted, and the abundance of marine organisms (including whales) has decreased drastically over the past century. Although these food webs are robust in the face of extreme seasonal change and have persisted in the face of long-term fluctuations over the years, tropical ecosystems are especially delicate in the face of human influences, especially commercial fishing, whaling, and pollution. These ecosystems are now severely suffering from a major crisis resulting from depleted fish stocks (Palomares and Pauly 2004; Fanning et al. 2007). While trying to gain an understanding of what could explain the collapse in their fisheries, many national research agencies of both Northwest African and Caribbean countries have been working at collecting data on the different species or links or these food webs. At the same time, some foreign agencies explain the crisis by proposing that whales are consuming fish and that this is why there is nothing left to catch in tropical areas. By doing so, they try to end the moratorium on commercial whaling at the International Whaling Commission (IWC) and resume the hunt of the so-far-protected whales. While many local managers and government members consider this to be a credible argument, most scientists assert that whales are not having much of an impact in these tropical breeding areas. In reality, it is now well- documented that overfishing is happening on a global scale (Jackson et al. 2001; Pauly et al. 2002; Baum et al. 2003; Myers and Worm 2003), and there is no scientific evidence for the existence of large-scale competition between marine mammals and fisheries (Kaschner 2004; Morissette 2007). Last but not least, given the importance of indirect effects in ecosystems, the true consequences of culling are unpredictable (Paine et al. 1998; Scheffer et al. 2001; Corkeron 2009). Nonetheless, ‘whales eat fish’ has become a very controversial issue and now comes back to the table year after year during IWC meetings (Gerber et al. 2009). Meanwhile, the public profile of “whales don’t eat fish”, promoted by many NGOs such as Lenfest Ocean Program, IFAW, WWF and Greenpeace, has never been higher.  a Cite as: Morissette, L. (2009) An ecosystem approach to the ‘whales eat fish’ issue. In: Morissette, L., Melgo, J.L., Kaschner, K. and Gerber, L.R. (eds.) Modelling the trophic role of marine mammals in tropical areas: data requirements, uncertainty, and validation. Fisheries Centre Research Reports 17(2). Fisheries Centre, University of British Columbia, Vancouver, Canada, pp. 3-5.  An ecosystem approach to the ‘whales eat fish’ issue, L. Morissette  4 It is now clear that ocean ecosystems throughout the world have experienced a dramatic shift in structure as a result of extensive fishing activities and the removal of top predators (see Estes et al. 2007). In order to address issues on the interactions between marine mammals and fisheries (such as the ‘whales eat fish’ issue), an ecosystem approach is essential, as there are a large number of indirect and direct interactions through which these two groups might influence each other (Bax 1998; Morissette et al. 2006). When these complex trophic interactions are taken into account, it has been shown in a number of cases that culling of marine mammals would not necessarily lead to recovery of fish stocks, nor otherwise benefit the commercial fishery (Punt and Butterworth 1995; Plagányi and Butterworth 2002; Morissette 2007; Gerber et al. 2009). Beneficial predations effects, e.g., marine mammals and other high-level predators increasing fisheries catches by feeding on other species that could be competing with the fleets, is more and more documented (Punt and Butterworth 1995, Walters and Kitchell 2001; Morissette). Only by considering all possible direct and indirect trophic linkages can the effect on current fisheries yields of the partial or complete removal of large whales be reliably assessed. Here, we examine the scientific evidence for the assertion that commercial fisheries are negatively impacted by whales. We use the Ecopath with Ecosim (EwE) approach, a quantitative whole-ecosystem model that tracks trophic flows in the food web from plankton, through pelagic and benthic fishes, all the way to marine mammals. EwE is one of several ecosystem modelling approaches that are widely used in understanding the interactions between marine mammals and fisheries (Morissette 2007). That approach is important because it represents a rational way of quantifying the trade-offs between sustainable exploitation of natural marine resources and conservation of charismatic fauna (Pitcher, 2005). The models also have the advantageous possibility of being tuned (or validated) to conventional stock assessment data or surveyed biomass estimates. The project focuses on two regions where the “whales eat fish” assertion has become a political and management issue: Northwest Africa and the Caribbean. We employ a precautionary approach to understanding management scenarios regarding the reduction of the abundance of great whales in Northwest Africa. While there is great uncertainty in many model parameters, we suggest that rather than engaging in discussions about the interactions between whales and fish in the absence of data, models can be developed with the best assumptions available and refined as more data become available (Hammill and Stenson 2007; Currie 2007). Models may also be used to consider the range of possible scenarios for a wide range of uncertainty about parameters. Addressing uncertainty is critical to providing useful management advice for the ‘whales eat fish’ issue, but presents a difficult challenge to whole-ecosystem modeling. Data on the biology, life-cycle, and exploitation of marine organisms that need to be captured in whole ecosystem simulations are often lacking for tropical ecosystems. Consequently, an in-depth uncertainty analysis has to be performed in both models to examine the extent to which our analyses are influenced by data quality and uncertainty. This report also documents two workshops, held in Dakar, Senegal on 15-17 May 2008 and in Bridgetown, Barbados on 25-27 September 2008 and through which the ‘whales eat fish’ and data availability issues were explored. Thus, this report documents the updated models following these workshops, where African and Caribbean researchers worked with our team on refining our models’ structure and data. The present report is freely available at the website of the Fisheries Centre of the University of British Columbia. ( www.fisheries.ubc.ca/publications/reports/fcrr.php ).  REFERENCES  Anonymous 2001a. What can we do for the coming food crisis in the 21st century? Institute of Cetacean Research. 4pp. Anonymous. 2001b. Increasing competition between fisheries and whales. Japan's whale research in the Western North Pacific (JARPA II). Fisheries Agency. Baum, J.K., Myers, R.A., Kehler, D.G., Worm, B., Harley, J. and Doherty, P.A.. 2003. Collapse and conservation on shark populations in the Northwest Atlantic. Science 299: 389-392. Bax, N.J. 1998. The significance and prediction of predation in marine fisheries. ICES Journal of Marine Sciences 55: 997-1030. Corkeron, P. 2009. Marine mammals’ influence on ecosystem processes affecting fisheries in the Barents Sea is trivial. Biology Letters 1098/rsbl.2008.0628 Currie, D. 2007. Whales, Sustainability and International Environmental Governance. Review of European Community & International Environmental Law 16: 45-57 Estes, J.A., DeMaster, D.P., Doak, D.F., Williams, T.M. and Brownell Jr., R.L. 2007. Whales, whaling, and ocean ecosystems. University of California Press, 418 p. Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  5 Fanning, L., Mahon, R. McConney, P., Angulo, J, Burrows, F., Chakalall, B., Gil, D., Haughton, M., Heileman, S., Martínez, Ostine, L., Oviedo, A., Parsons, S., Phillips T.,  Arroya, C.S., Simmons, B., and C. Toro. A large marine ecosystem governance framework. Marine Policy 31: 434-443. Gerber, L., Morissette, L., Kaschner, K., and Pauly, D. 2009. Should whales be culled to increase fishery yield? Science 323: 880-881. Hammill, M.O. and Stenson, G.B. 2007. Application of the precautionary approach and conservation reference points to management of Atlantic seals. ICES Journal of Marine Science 64: 702–706. Holt, S.J. 2007. Whaling: Will the Phoenix rise again? Marine Pollution Bulletin 54: 1081-1086 Jackson, J. B. C., M.X. Kirby, W. H. Berger, K.A: Bjorndal, L. W. Botsford, B. J. Bourque, R. H. Bradbury, R. Cooke, J. Erlandson, J. A. Estes, T. P. Hughes, S. Kidwell, C. B. Lange, H. S. Lenihan, J. M. Pondolfi, C. H. Peterson, R. S. Steneck, M. J. Tegner, and R. R. Warner. 2001. Historical overfishing and the recent collapse of coastal ecosystems. Science 293: 629-638. Jackson, J.B.C. 2007. When ecological pyramids were upside down. pp. 23-37 In Estes, J.A., DeMaster, D.P., Doak, D.F., Williams, T.M. and Brownell Jr., R.S. (Eds.) Whales, Whaling, and Ocean Ecosystems. University of California Press, Berkeley and Los Angeles, California. Kaschner, K. 2004. Modelling and mapping of resource overlap between marine mammals and fisheries on a global scale. Ph.D., University of British Columbia. Katona, S. and Whitehead, H. 1988. Are cetaceans ecologically important? Oceanography and Marine Biology Annual Reviews 26: 553-568. Komatsu, M. and Misaki, S. 2003. Whales and the Japanese: how we have come to live in harmony with the bounty of the sea. The Institute of Cetacean Research, Tokyo, 170 pp. Morissette, L. 2007. Complexity, cost and quality of ecosystem models and their impact on resilience: a comparative analysis, with emphasis on marine mammals and the Gulf of St. Lawrence. PhD thesis, Zoology, University of British Columbia, Vancouver BC, Canada. Myers, R. A. and Worm, B. 2003. Rapid worldwide depletion of predatory fish communities. Nature 423: 280-283. Paine, R.T., Tegner, M.J., and Johnson, E.A. 1998. Compounded perturbations yield ecological surprises. Ecosystems 1: 535-545. Palomares, M.L.D. and Pauly, D. (Eds.) 2004. West African marine ecosystems : models and fisheries impacts. Fisheries Centre Research Reports 12(7). Pauly, D., Christensen, V., Guénette, S. Pitcher, T.J., Sumaila, U.R., Walters, C.J., Watson, R. and Zeller, D. 2002. Towards sustainability in world fisheries. Nature 418: 689-695. Pitcher, T.J. (2005) Simulating Antarctic ecosystems and fisheries: weapons of mass construction. In: Palomares, M.L.D., Plagányi, E.E. and D.S. Butterworth. 2002. Competition with fisheries. pp 268-273 In Perrin, W.F., Würsig, B., and Thewissen, H.G.M. (Eds.) Encyclopedia of Marine Mammals. Academic Press, San Diego. Pruvost, P., Pitcher, T.J., Pauly, D. (eds.) Modeling Antarctic marine ecosystems. Punt, A.E. and Butterworth, D.S. 1995. The effects of future consumption by the Cape fur seal on catches and catch rates of the Cape hakes. 4. Modelling the biological interaction between Cape fur seals Arctocephalus pusillus pusillus and the Cape hakes Merluccius capensis and M. paradoxus. South African Journal of Marine Science 16: 255-285. Scheffer, M., S. Carpenter, J. A., Foley, C. Folke, and Walker, B. 2001. Catastrophic shifts in ecosystems. Nature 413: 591-596. Springer, A.M., Estes, J.A., van Vliet, G.B., Williams, T.M., Doak, D.F., Danner, E.M., Forney, K.A., and Pfister, B. 2003. Sequential megafaunal collapse in the North Pacific Ocean; an ongoing legacy of industrial whaling? Proceedings of the National Academy of Sciences 100: 12 223-12 228. Struck, D. 2001. Japan blames whales for lower fish catch. International Herald Tribune (July 28-29). Tamura, T. 2003. Regional assessments of prey consumption and competition by marine cetaceans in the world. In Responsible Fisheries in the Marine Ecosystem. pp. 143-170 In Sinclair, M. and Valdimarsson, G. (Eds). Responsible Fisheries in the Marine Ecosystem, Fishery Industries Division, CAB International: FAO, Rome (Italy) and Wallingford (United Kingdom). Walters, C. and J.F. Kitchell. 2001. Cultivation/Depensation Effects on Juvenile Survival and Recruitment: Implications for the Theory of Fishing. Canadian Journal of Fisheries and Aquatic Sciences 58: 39-50.   Food web models and data for NW African ecosystem, L. Morissette et al.  6 FOOD WEB MODEL AND DATA FOR STUDYING THE INTERACTIONS BETWEEN MARINE MAMMALS AND FISHERIES IN THE NORTHWEST AFRICAN ECOSYSTEMa Lyne Morissette1,2,3, Jenny Lynn Melgo1, Kristin Kaschner4, Leah Gerber1, Idrissa Lamine Bamy5,  1Arizona State University, School of Life Sciences, P.O. Box 874501, Tempe, AZ 85287-4501, USA 2Institut des sciences de la mer de Rimouski, 310, Allée des Ursulines, C.P. 3300, Rimouski, QC, G5L 3A1, CANADA 3Fisheries Centre, The University of British Columbia 2202 Main Mall, Vancouver BC V6T 1Z4, CANADA 4Albert-Ludwigs-University, Institute of Biology I (Zoology) Evolutionary Biology & Ecology Lab, Freiburg, GERMANY 5Centre National des Sciences Halieutiques de Boussoura, BP-3738 Boussoura, Conakry, GUINEA Lyne.Morissette@globetrotter.net Jennylynn.Melgo@asu.edu Kristin.Kaschner@biologie.uni-freiburg.de Leah.Gerber@asu.edu ibamy@caramail.com  ABSTRACT This report describes the data and methodology used to construct a model for Northwest Africa during the late 1980s. The model for Northwest Africa includes the Large Marine Ecosystem (LME) of the Canary Current, which is located on the eastern part of the Atlantic Ocean, and bounded by the coasts of Morocco, Mauritania, Senegal, Guinea-Bissau, the Canary Islands (Spain), Gambia, Cape Verde and Western Sahara. The model was developed to examine the trophic interactions between marine mammals and fisheries and uses simulations to examine the potential impact of a reduction in the abundance of great whales on fishery yield. The model includes 10 marine mammals groups and 17 additional groups comprised of fish, seabirds, invertebrates, and plankton. Both local and foreign fleets are also included in the model.  INTRODUCTION A mass-balanced model of the Northwest coast of Africa was constructed using EwE (Christensen and Walters 2004). It took as a starting point an Ecopath model developed by Samb and Mendy (2004), as well as six other Ecopath models developed in the context of “Fisheries Information and Analysis System”  a Cite as: Morissette, L., Melgo, J.L., Kaschner, K., Gerber, L., and Bamy, I.L. (2009) Food web model and data for studying the interactions between marine mammals and fisheries in the Northwest African ecosystem. In: Morissette, L., Melgo, J.L., Kaschner, K. and Gerber, L.R. (eds.) Modelling the trophic role of marine mammals in tropical areas: data requirements, uncertainty, and validation. Fisheries Centre Research Reports 17(2). Fisheries Centre, University of British Columbia, Vancouver, Canada, pp. 6-47. Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  7 (SIAP), after two workshops in Dakar (February and August 2001), which were attended by scientists from the six countries represented by the models: Cape Verde, Gambia, Guinea, Guinea-Bissau, Mauritania, and Senegal. We used the Senegambian ecosystem model (Samb and Mendy 2004) as a starting point, and included relevant information about marine mammals and other important trophic group for the greater Northwest Africa region. This report describes the data sources used for constructing the Ecopath with Ecosim model of Northwest Africa, and the uncertainty analyses performed on input and output data.  The aim of this report is to present our preliminary results to the participants of the “Whale and fish interactions: are great whales a threat to fisheries?” workshop in Dakar, Senegal on 8-9 May, 2008. Preliminary results of the research are still being refined and will be presented at the meeting. We hope to receive expert advice and feedback on our research methodology, data and preliminary results during the discussions held at the workshop. We also hope to confirm that we are not missing key data that should be included in our model.  The model will be continuously updated as more data are collected.  Ultimately, our model may be used as a basis to evaluate the trophic role of great whales in the ecosystem of Northwest Africa.  MATERIALS AND METHODS STUDY AREA Our study area is located off the coast of Northwest Africa, bounded by Morocco, Mauritania, Senegal, Guinea-Bissau, the Canary Islands (Spain), Gambia, Cape Verde and Western Sahara (Figure 1). This area defined by the United States National Oceanic and Atmospheric Administration (NOAA) as the Canary Current large marine ecosystem (LME), and is included in the Food and Agriculture Organization’s (FAO’s) Eastern Central Atlantic (Major Fishing Area 34), mainly covering subdivision 34.1 (Northern coastal), and part of subdivision 34.2 (Northern oceanic). The specific area of this model covers latitudes from 8.5 N to 35.97 N, and longitudes from 30 W to 6.5 W, for a total area of 3,561,029 km2 (Figure 1). Our study area includes the continental shelf as well as the deeper area. The Northwest coast of Africa is characterized by the presence of the Canary Current, which flows along the African coast from north to south between 30°N and 10°N and offshore to 20°W (Fedoseev 1970). As a consequence, one major characteristic of this ecosystem is that it shows a major upwelling and other seasonal nutrient enrichments. Climate is the primary force driving the dynamics of this ecosystem, with intensive fishing as the secondary driving force (Bas 1993).  Food web models and data for NW African ecosystem, L. Morissette et al.  8 Figure 1. Map of the study area; the coast of Northwest Africa.  HARVESTING ON THE ECOSYSTEM Whaling Whaling is known to occur in Northwest African waters (Reeves 2002), but to our knowledge no official data on actual harvests are available. The only record available documents aboriginal subsistence whaling in Equatorial Guinea, where Indigenous Africans principally target humpback whales (and mainly calves). No estimate is available on the annual take, but Aguilar (1985) guessed that it was approximately three humpback whales annually. These whales are assumed to be part of a Southern Hemisphere population of humpback whales (Reeves 2002). Fishery The area of Northwest Africa is generally thought to be overexploited for most coastal demersal species (Samb and Mendy 2004). As a result, fishing effort is often reported for species such as sardinellas, for which acoustic survey shows a relatively stable abundance. However, Samb and Mendy (2004) mention that this shift from fishing coastal demersal species to a situation where sardinellas are very important in the catch should be better addressed since sardinellas have an important role in maintaining the structure of the food web. The Northwest African ecosystem supports both local and foreign fleets, representing an average of 2,153,091 tonnes of fish caught annually in the system. From 1987 to 2004, local fleets generally caught the majority of this biomass, but the proportion of local fleets vs. foreign fleets varied from a minimum of 53.1% of catches taken by local fleets in 1990 to a maximum of 81.7% of the catches in 1994 (Figure 2). Demersal fisheries have increased substantially over the last few decades (Gascuel et al. 2007), but it seems that few studies can describe catches and fishing effort adequately, and that FAO data are not reliable for the area (Gascuel et al. 2007). Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  9    Local fleets Figure 2. Local and foreign catches off the coast of Northwest Africa from 1987-2004. Source: Sea Around Us Project database, www.searoundus.org Foreign fleets Food web models and data for NW African ecosystem, L. Morissette et al.  10 We synthesized available information about the catch composition for the period 1987-2004 by species, year, and country for local and foreign fleets fishing off the coast of Morocco, Mauritania, Senegal, Guinea-Bissau, Canary Islands, Gambia, Cape Verde, and Western Sahara (Tables 1 and 2).  We relied on estimates of unreported catch and by-catch of industrial fisheries and the database of fisheries catch developed by the Sea Around Us Project (SAUP) at the University of British Columbia (www.seaaroundus.org). Fisheries catch taken in all three study areas were obtained from the Sea Around Us database (Sea Around Us, 2008. A global database on marine fisheries and ecosystems. World Wide Web site www.seaaroundus.org. Fisheries Centre, University British Columbia, Vancouver (British Columbia, Canada). [Visited February 2008]). Time series of annual total catches taken between 1987 and 2004 were specified by the respective countries fishing in the area and by the taxa that were taken. We assigned taxa to the appropriate functional group of each Ecopath model using available information about life history, ecology and habitat preferences of the taxa. Catches were then divided into local and foreign fisheries. Local fisheries were defined as all countries bordering on our study area, disregardful of whether or not catches were taken within each countries own EEZ waters or in neighboring waters. All other fishing countries were defined as foreign fleets (Figure 2).  We used our database to reconstruct the catch composition for the period 1987-2004.  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  11 Table 1. Time series of local fleets’ catches by trophic group in the Ecopath model for Northwest Africa (in ‘1000 tonnes). Year 10  Large pelagics 11  Mesopelagic predators 12 Bathy- demersal predators 13   Sharks 14   Rays 15  Coastal tunas 16  Coastal demersals 17   Clupeids 18 Other coastal pelagics 19  Cephal o-pods 20  Crusta- ceans 21   Benthos Total 1987 26.915 5.388 24.568 6.017 1.774 10.672 120.415 500.891 213.504 134.650 20.709 6.845 1072.348 1988 23.375 3.734 21.582 6.521 2.427 9.162 107.468 629.245 210.107 136.912 23.308 4.702 1178.544 1989 23.735 3.972 24.713 6.037 2.208 6.745 109.722 672.237 147.371 158.504 30.032 3.401 1188.676 1990 30.165 4.622 24.757 5.454 3.807 6.166 109.685 667.183 169.282 169.978 20.981 3.453 1215.534 1991 32.851 2.534 24.264 3.217 3.370 3.613 109.248 717.077 161.821 206.445 24.976 4.914 1294.329 1992 27.097 2.544 27.128 3.471 4.001 3.972 115.562 679.073 188.770 165.040 20.254 4.595 1241.507 1993 27.447 3.775 25.942 3.215 4.134 3.801 115.915 761.000 196.992 174.518 22.201 5.249 1344.189 1994 25.501 3.272 27.374 4.211 5.477 3.553 123.068 854.852 194.105 166.253 24.227 5.745 1437.637 1995 30.396 3.507 28.191 5.936 6.794 5.145 132.434 942.914 187.792 171.359 27.164 8.681 1550.315 1996 24.826 7.201 28.259 10.190 7.115 7.404 141.367 863.837 157.882 173.034 35.566 8.366 1465.047 1997 28.102 4.159 23.436 21.334 13.816 8.647 147.355 987.874 216.845 131.279 27.518 6.743 1617.110 1998 25.795 4.560 19.726 11.751 9.815 8.320 140.712 931.278 197.237 171.030 42.228 10.613 1573.065 1999 25.745 4.472 21.831 12.194 8.540 5.336 136.188 841.400 186.154 217.558 32.879 7.465 1499.761 2000 22.495 4.246 16.535 14.592 10.27 8 6.836 152.896 927.665 214.420 199.105 29.896 7.165 1606.129 2001 25.208 4.149 23.614 15.614 9.234 6.820 156.339 1171.271 225.264 189.168 31.663 7.841 1866.186 2002 21.883 3.893 19.642 12.537 7.749 6.291 150.478 1050.218 243.100 88.259 19.687 6.330 1630.067 2003 26.922 3.701 24.948 10.870 6.629 5.244 165.718 1078.499 230.319 85.690 20.898 6.215 1665.654 2004 21.440 3.747 25.759 4.318 13.93 0 4.607 152.834 1047.249 294.973 62.420 17.416 10.185 1658.877  Food web models and data for NW African ecosystem, L. Morissette et al.  12 Table 2. Time series of foreign fleets’ catches by trophic group in the Ecopath model for Northwest Africa (in ‘1000 tonnes). Year 10  Large pelagics 11  Mesopelagic predators 12  Bathydemersal predators 13   Sharks 14   Rays 15  Coastal tunas 16  Coastal demersals 17   Clupeids 18 Other coastal pelagics 19   Cephalopods 20   Crustaceans 21   Benthos Total 1987 18.133 0.000 3.537 0.860 1.688 10.972 44.027 314.194 344.466 23.178 8.488 0.426 769.969 1988 23.423 0.000 4.339 0.919 2.416 9.774 37.594 359.614 445.571 19.852 6.710 0.140 910.352 1989 24.700 0.000 3.028 0.862 2.077 10.973 30.130 565.703 327.247 24.036 5.943 0.325 995.025 1990 29.010 0.000 2.071 0.903 2.208 10.022 41.111 635.854 333.376 13.816 5.261 0.029 1073.660 1991 27.706 0.000 1.465 1.212 2.365 6.323 52.960 615.550 375.413 14.854 5.275 0.000 1103.124 1992 23.680 0.000 0.783 1.796 3.810 1.941 50.265 268.615 277.797 11.726 4.730 0.000 645.142 1993 30.062 0.000 0.064 1.491 4.674 4.117 44.020 117.591 202.228 14.379 6.365 0.667 425.658 1994 29.583 0.000 0.259 1.408 4.840 0.569 41.305 64.681 165.995 8.650 4.089 0.000 321.378 1995 27.854 0.001 0.464 0.139 2.664 0.490 20.222 96.962 375.167 15.962 4.197 0.040 544.160 1996 26.786 0.001 0.149 0.104 1.840 6.720 17.739 214.901 285.468 11.998 3.219 0.000 568.925 1997 22.050 0.000 0.647 0.037 0.963 7.052 27.036 237.487 276.764 17.916 3.969 0.010 593.931 1998 26.707 0.000 2.863 0.072 0.310 6.194 26.508 373.057 297.179 20.290 4.173 0.009 757.360 1999 26.917 0.000 0.898 0.080 0.433 6.872 27.587 340.364 277.016 32.015 3.861 0.005 716.048 2000 21.277 0.000 1.863 0.105 0.732 6.392 31.781 299.077 330.533 17.193 3.322 0.003 712.278 2001 25.377 0.000 1.555 0.427 0.683 3.177 39.891 370.917 393.755 23.882 3.851 0.003 863.518 2002 19.030 0.000 2.311 0.338 0.507 4.217 20.634 302.658 276.657 3.262 1.089 0.000 630.702 2003 21.486 0.000 2.001 0.433 0.625 2.753 23.338 301.637 177.633 4.339 1.198 0.000 535.443 2004 23.157 0.000 1.109 0.016 0.582 2.463 20.238 263.034 168.238 4.295 0.865 0.000 483.996  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  13   MODEL DESCRIPTION The ecosystem model for Northwest Africa was modified based on a model for the Senegambian ecosystem in the 1990s (Samb and Mendy 2004). The Senegambian ecosystem model was used because it was representative of a typical ecosystem in Northwest Africa, and could thus be extrapolated to represent our larger study area. This was not the case, for example, for the Mauritanian EEZ model (Sidi and Guénette 2004), which represents an upwelling ecosystem, with more particular characteristics (D. Gascuel pers. comm.). Originally including 18 trophic groups, the model was modified to include a better representation of the different marine mammals species, and also commercially important fish groups for the Northwest African region. In order to construct comparable models for other focal areas where we are studying the interaction between marine mammals and fisheries (e.g. Caribbean and South Pacific), biomass was  aggregated into distinct functional feeding groups.  These groups were defined based on the similarities in food habits, habitats, and biological variables (Essington 2006). The model for Northwest Africa presented here consists of 27 trophic groups: 10 marine mammal groups, one seabirds group, 9 fish groups, cephalopods, large crustaceans, 2 benthos groups, zooplankton, phytoplankton, and detritus (Table 3). For most trophic groups, we used adult diet data, as most of the biomass estimates available were only for the spawning stock or adult biomass. Cetacean groups were included in this ecosystem model at a high taxonomic resolution because of the aim of our study. It is important to note though, that feeding patterns of the baleen whales in particular are such, that very little of their annual food consumption (if any) is taken in Northwest African waters. The tropical, warmer waters covered by our three study areas represent, in fact, breeding areas for baleen whales, in which these species are known to mainly fast (Brodie 1975; Sergeant 1977; Brown and Lockyer 1984; Corkeron and Connor 1999; Perry et al. 1999, Clapham 2002; Jann et al. 2003). Indeed, blue, fin, sei and humpback whales spend their feeding season in the colder waters of north Atlantic in the Atlantic part of Northern Hemisphere or in the Antarctic waters in the Southern Hemisphere. Lockyer (1981) showed that most baleen whales feed considerably less in their breeding grounds than they would in feeding areas. She proposed that the average amounts consumed off breeding areas probably amount to about 10% or less of that in the feeding ground. This reduced consumption rate of 10% was then applied to many other ecosystems by different authors (Brown and Lockyer 1984; Mohammed 2003). In our model, we treated these migratory species as part of the ‘system’ all the time in terms of some calculated impacts like fishing, whaling, and general trophic interactions.  To account for the fact that most feeding activity occurs outside the system, we set a high diet proportion as ‘Import’ in the Ecopath diet consumption matrix (Christensen et al. 2005). Food web models and data for NW African ecosystem, L. Morissette et al.  14  Table 3. List of trophic groups and species included in the Ecopath model for Northwest Africa. Species in bold represent to key species for each trophic group. Ecopath group Species 1. Minke whales Balaenoptera acutorostrata 2. Fin whales Balaenoptera physalus 3. Humpback whales Megaptera novaeangliae 4. Bryde’s whales Balaenoptera brydei 5. Sei whales Balaenoptera borealis 6. Blue whales Balaenoptera musculus 7. Sperm whales Physeter macrocephalus 8. Killer whales Orcinus orca 9. Beaked whales Mesoplodon densirostris, M. europaeus, Ziphius caviostris 10. Small cetaceans Delphinus delphis, Feresa attenuate, Globicephala macrorhynchus, Grampus griseus, Kogia breviceps, Kogia simus, Lagenodelphis hosei, Peponocephala electra, Pseudorca crassidens, Sousa teuszii, Stenella attenuate, Stenella clymene, Stenella coeruleoalba, Stenella frontalis, Stenella longisrostris, Steno bredanensis, Tursiops truncatus 11. Seabirds Actitis hypoleucos, Calidris ferruginea, Calonectris diomedea, Ceryle rudis, Chlidonias niger, Halcyon malimbica, Limosa lapponica, Numenius phaeopus, Oceanites oceanicus, Pagrodama nivea, Pelecanus rufescens, Phalacrocorax africanus, Phoenicopterus rubber, Pluvialis squatarola, Sterna caspia, Sterna hirundo 12. Large pelagics Acanthocybium solandri, Brama brama, Centrolophidae, Coryphaena hippurus, Cubiceps gracilis, Istiophoridae, Istiophorus albicans, Istiophorus platypterus, Katsuwonus pelamis, Makaira nigricans, Ranzania laevis, Ruvettus pretiosus, Schedophilus medusophagus, Tetrapturus albidus, Tetrapturus pfluegeri, Thunnus alalunga, Thunnus albacares, Thunnus obesus, Thunnus thynnus, Xiphias gladius 13. Mesopelagics predators Aphanopus carbo, Astronesthes niger, Atherina presbyter, Benthosema glaciale, Borostomias elucens, Chauliodus danae, Diplospinus multistriatus, Evermannella balbo, Lampris guttatus, Lepidocybium flavobrunneum, Leptostomias gladiator, Maurolicus muelleri, Micromesistius poutassou, Micromesistius poutassou, Mora moro, Moridae, Myctophum asperum, Myctophum nitidulum, Myctophum punctatum, Nealotus tripes, Photonectes margarita, Polyacanthonotus challengeri, Rhadinesthes decimus, Sternoptyx diaphana, Stomias boa boa, Stomiidae, Trachichthyidae, Trachyrincus scabrus, Vinciguerria nimbaria, Xenodermichthys copei 14. Bathydemersal predators Beryx sp., Beryx decadactylus, Caelorinchus caelorhincus caelorhincus, Caproidae, Chimaera monstrosa, Coryphaenoides rupestris, Coryphaenoides zaniophorus, Gadiformes, Gempylidae, Helicolenus dactylopterus dactylopterus, Lophiidae, Lophius budegassa, Lophius piscatorius, Lophius vaillanti, Lotidae, Merlucciidae, Merluccius merluccius, Merluccius polli, Merluccius senegalensis, Muraena helena, Nezumia aequalis, Nezumia sclerorhynchus, Phycidae, Phycis blennoides, Phycis phycis, Polyprion americanus, Pristis pectinata, Spectrunculus grandis, Synaphobranchus kaupii 15. Sharks Alopias sp., Alopias superciliosus, Alopias vulpinus, Alopiidae, Carcharhinidae, Carcharhinus falciformis, Carcharhinus limbatus, Carcharhinus longimanus, Carcharhinus obscurus, Carcharhinus plumbeus, Centrolophidae, Centrophorus granulosus, Centroscyllium fabricii, Centrophorus squamosus, Centrophorus uyato, Centroscymnus coelolepis, Centroscymnus cryptacanthus, Centroscymnus crepidater, Cetorhinus maximus, Dalatias licha, Deania calcea, Elasmobranchii, Etmopteridae, Etmopterus princes, Etmopterus pusillus, Galeorhinus galeus, Galeus melastomus, Galeus polli, Ginglymostoma cirratum, Hexanchus griseus, Isurus sp., Isurus oxyrinchus, Lamna nasus, Lamnidae, Mustelus asterias, Mustelus mustelus, Prionace glauca, Pristidae, Rhizoprionodon acutus, Scyliorhinidae, Scyliorhinus canicula, Scyliorhinus stellaris, Sphyrna lewini, Sphyrna zygaena, Sphyrnidae, Squalidae, Squalus acanthias, Squalus blainville, Squalus megalops, Squatina squatina, Squatinidae, Triakidae Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  15 16. Rays Dasyatidae, Dasyatis margarita, Dasyatis pastinaca, Dipturus batis, Dipturus oxyrinchus, Gymnura altavela, Leucoraja naevus, Myliobatidae, Myliobatis aquila, Raja clavata, Raja miraletus, Raja montagui, Raja straeleni, Rajidae, Rajiformes, Rhinobatidae, Rhinobatos cemiculus, Rhinobatos rhinobatos, Rhinoptera bonasus, Rhinoptera marginata, Torpedinidae, Torpedo sp. 17. Costal tunas Auxis rochei, Auxis thazard, Euthynnus alletteratus, Orcynopsis unicolor, Sarda sarda,Scomberomorus tritor 18. Coastal demersals Acanthuridae, Albula vulpes, Ammodytidae, Anthias anthias, Aphia minuta, Apogon imberbis, Apogonidae, Argentina sphyraena, Argyrosomus regius, Ariomma bondi, Ariidae, Ariomma melanum, Arius heudelotii, Arnoglossus laterna, Aulopus cadenati, Balistidae, Boops boops, Bothidae, Bothus podas, Brachydeuterus auritus, Brotula barbata, Campogramma glaycos, Capros aper, Cepola macrophthalma, Chaetodon hoefleri, Charis charis, Chelidonichthys obscurus, Chlorophthalmus agassizi, Conger conger, Congridae, Ctenolabrus rupestris, Cynoglossidae, Cynoglossus senegalensis, Dentex  angolensis, Dentex  canariensis, Dentex  dentex, Dentex  gibbosus, Dentex  macrophthalmus, Dentex  maroccanus, Dicentrarchus sp.,Dicentrarchus labrax, Dicologlossa cuneata, Diplodus bellottii, Diplodus cervinus cervinus, Diplodus sargus cadenati, Diplodus vulgaris, Drepane africana, Echeneidae, Emmelichthyidae, Epinephelus  marginatus, Epinephelus aeneus, Epinephelus goreensis, Eucinostomus melanopterus, Fistularia tabacaria, Gaidropsarus sp., Galeoides sp., Galeoides decadactylus, Gerres nigri, Gobius niger, Gobius paganellus, Haemulidae, Halobatrachus didactylus, Labridae, Lepidorhombus sp., Lepidotrigla cadmani, Lepidotrigla dieuzeidei, Lethrinus atlanticus, Lithognathus mormyrus, Liza aurata, Liza dumerili,  Liza falcipinnis, Liza grandisquamis, Liza ramado, Lutjanus sp., Lutjanus goreensis, Macroramphosus scolopax, Molva sp., Microchirus sp., Microchirus boscanion, Microchirus variegates, Monochirus hispidus, Mugilidae,  Mugil capurrii, Mugil cephalus, Mullidae, Mullus  barbatus, Mullus sp., Mullus surmuletus, Muraenidae, Mycteroperca rubra, Oblada melanura, Plectorhinchus macrolepis, Pagrus sp., Pagrus pagrus, Pagrus caeruleostictus, Pagellus sp., Pagellus bellottii bellottii, Pagellus acarne, Pagellus bogaraveo, Pagellus erythrinus, Pegusa lascaris, Pentanemus quinquarius, Platichthys flesus, Plectorhinchus macrolepis, Plectorhinchus mediterraneus, Pleuronectidae, Pleuronectiformes, Pleuronectes platessus, Polynemidae, Pomacentridae, Polydactylus quadrifilis, Pontinus kuhlii, Pomadasys jubelini, Pomadasys incisus, Pomadasys perotaei, Pomadasys rogerii, Pseudotolithus typus, Pseudotolithus senegalensis, Pseudotolithus elongatus, Pseudotolithus senegallus, Pseudupeneus prayensis, Pteroscion peli, Pseudupeneus prayensis, Rachycentron canadum, Sarpa salpa, Saurida brasiliensis, Scaridae, Sciaenidae, Sciaena umbra, Schedophilus pemarco, Scophthalmidae, Scophthalmus rhombus, Serranidae, Selene dorsalis, Scorpaenidae, Scorpaena maderensis, Scorpaena notata, Syacium guineensis, Soleidae, Solea senegalensis, Solea solea, Sparidae,  Sparus auratus,  Sparus caeruleostictus, Spondyliosoma cantharus, Stephanolepis hispidus, Stromateus fiatola, Symphodus mediterraneus, Symphodus melops, Synagrops microlepis, Synaptura lusitanica lusitanica, Trachinus draco, Trachinocephalus myops, Triglidae, Tetraodontidae, Trisopterus minutus, Trisopterus luscus, Umbrina cirrosa, Umbrina canariensis, Zeus faber, Zenopsis conchifer 19. Clupeids Alosa alosa, Clupeidae, Clupeiformes, Engraulis encrasicolus, Ethmalosa fimbriata, Ilisha Africana, Sardina pilchardus, Sardinella sp., Sardinella aurita, Sardinella maderensis, Sprattus sprattus 20. Other coastal pelagics Alectis alexandrinus, Aphanopus intermedius, Belonidae, Carangidae, Caranx sp., Caranx hippos, Caranx rhonchus, Caranx senegallus, Cheilopogon heterurus, Chloroscombrus chrysurus, Decapterus sp., Decapterus punctatus, Dicentrarchus punctatus, Elops lacerta, Exocoetidae, Exocoetus obtusirostris, Hemiramphidae, Hemiramphus sp., Lepidopus caudatus, Lichia amia, Pomatomus saltatrix, Promethichthys pometheus, Regalecus glesne, Scomber sp., Scomber japonicus, Scomber scombrus, Scomberesox saurus saurus, Scomberomorus sp., Scombridae, Seriola sp., Sphyraena sp., Sphyraena barracuda, Spicara sp., Trachinotus sp., Trachinotus ovatus, Trachurus sp., Trachurus mediterraneus, Trachurus picturatus, Trachurus trachurus, Trachurus trecae, Trichiuridae, Trichiurus lepturus, Tylosurus acus acus 21. Cephalopods Alloteuthis subulata, Cephalopoda, Illex coindetii, Loliginidae, Loligo sp., Loligo vulgaris, Octopodidae, Octopus vulgaris, Ommastrephidae, Sepia bertheloti, Sepia elobyana, Sepia officinalis, Sepia orbignyana, Sepiidae, Teuthida, Todarodes sagittatus Food web models and data for NW African ecosystem, L. Morissette et al.  16 22. Crustaceans Aristeidae, Aristeus antennatus, Aristeus varidens, Brachyura, Calappa rubroguttata, Cancer pagurus, Carcinus maenas, Crangon sp., Crangon crangon, Crangonidae, Geryon sp., Geryon maritae, Homarus gammarus, Leucosiidae, Maja squinado, Metapenaeus, Munidae, Natantian decapods, Necora puber, Nephrops norvegicus, Paguridae, Palaemonidae, Palinurus sp., Palinurus elephas, Palinurus mauritanicus, Panulirus regius, Panulirus sp., Parapenaeopsis sp., Parapenaeopsis atlantica, Parapanaeus longirostris, Penaeidae, Penaeus sp., Penaeus kerathurus, Penaeus notialis, Pleoticus robustus, Plesionika heterocarpus, Plesiopenaeus edwardsianus, Portunidae, Scyllaridae, 23. Benthos Anthozoa, Arca sp., Arcidae, Bivalvia, Cardiidae, Cardium edule, Chama crenulata, Conidae, Crassostrea sp., Crepidula porcellana, Cymbium sp, Donacidae, Donax sp., Epizoanthidae, Gastropoda, Glycymerididae, Haliotidae, Haliotis tuberculata, Modiolus sp., Murex sp., Muricidae, Mytilidae, Naticidae, Ostrea edulis, Patella sp., Pecten maximus, Pectinidae, Porifera, Pyura dura, Ruditapes decussates, Solen sp., Solenidae, Tapes sp., Thais haemastoma, Veneridae, Venus rosalina, Venus verrucosa, Volutidae 24. Benthic producers Algae, benthic bacteria 25. Zooplankton Copepoda, Hydrozoa, Scyphozoa, fish larvae, eggs 26. Phytoplankton 27. Detritus  Groups 1 - 10. Marine mammals Ten groups of marine mammals were added to the original model for the purpose of our study. We considered species of commercial interest individually and aggregated remaining species to facilitate model simulations. Given the lack of local, long-term dedicated surveys to provide reliable cetacean abundance estimates, density estimates had to be derived from a global database (Kaschner 2004). However, comparison with other densities from surveys conducted in similar habitats are ground- truthing these estimates (Table 4). Similarly, most estimates about diet for marine mammals came from other areas. The lack of studies on the diet of marine mammals off the coast of Northwest Africa is mainly explained by the fact that these stocks of whales are not known to feed in this area (Brodie 1975; Sergeant 1977; Lockyer 1981; Brown and Lockyer 1984; Corkeron and Connor 1999; Perry et al. 1999, Clapham 2002; Jann et al. 2003).  When data were lacking for diet for marine mammals in Northwest Africa, we relied on diet information from feeding areas. This approach has been employed in other ecosystem models (Bundy et al. 2000; Okey 2001; Guénette and Christensen 2005; Morissette 2007). Nonetheless, our results should be interpreted with caution given the uncertainty in our diet parameters.  To address this issue, we performed different model scenarios by changing diet composition to increase the proportion of commercially important fish in order to examine the extent to which our diet assumptions might change the outcomes of our models.  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  17 Table 4. Comparison of predicted cetacean densities in Northwest African waters based on global model developed by Kaschner et al (2006) and Kaschner (2004) and observed densities in similar habitats (subtropical & tropical waters). A = aerial surveys, S = ship based surveys.  Density estimates that are corrected for animals missed on the track-line are indicated in the G(0) corrected column. All other observed estimates might represent underestimations. Bold observed density values represent lowest and highest observed estimates, respectively for each species. Common Name Estimated density [animals / 1000 km2] Observed density [animals / 1000 km2] CV G(0) corrected Geographic area Survey years Survey type Source Blue whale 0.03 0.07 0.24 no Eastern Tropical Pacific 1986-1990 S Wade and Gerrodette 1993 Blue whale 0.03 1.10 0.33 yes NE Pacific, Baja California 1993 S Calambokidis and Barlow 2004 Blue whale 0.03 0.00 0.00 yes NE Pacific, west coast US 1996 S Calambokidis and Barlow 2004 Blue whale 0.03 3.11 0.28 yes NE Pacific, west coast US 1996 S Barlow 2003a Blue whale 0.03 0.95 0.44 yes NE Pacific, west coast US 2001 S Barlow 2003a Blue whale 0.03 0.11 0.99 yes NE Pacific, California inshore 1991-1992 A Forney et al. 1995 Blue whale 0.03 3.27 0.24 yes NE Pacific, west coast US 1991-1993 S Barlow 2003a Blue whale 0.03 1.26 0.27 yes NE Pacific, California offshore 1991-1996 S Calambokidis and Barlow 2004 Blue whale 0.03 4.96 0.13 yes NE Pacific, California inshore 1991-1996 S Calambokidis and Barlow 2004 Blue whale 0.03 0.76 0.50 no SW Indian Ocean, Madagaskar plateau (southern block) 1996 S Best et al. 2003 Blue whale 0.03 0.82 0.65 no SW Indian Ocean, Madagaskar plateau (northern block) 1996 S Best et al. 2003 Bryde's whale 0.49 0.11 0.61 no NW Atlantic, northern Gulf of Mexico (SEFSC) 1996-2001 S Mullin and Fulling 2004 Bryde's whale 0.49 0.67 0.20 no Eastern Tropical Pacific 1986-1990 S Wade and Gerrodette 1993 Bryde's whale 0.49 0.67 0.21 no Eastern Tropical Pacific 1998 S Gerrodette and Forcada 2002 Bryde's whale 0.49 0.50 0.24 no Eastern Tropical Pacific 1999 S Gerrodette and Forcada 2002 Bryde's whale 0.49 0.48 0.20 no Eastern Tropical Pacific 2000 S Gerrodette and Forcada 2002 Bryde's whale 0.49 0.19 0.45 yes NE Pacific, Hawaiin waters 2002 S Barlow 2006 Bryde's whale 0.49 0.06 0.53 yes NE Pacific, west coast US 1991-1993 S Barlow 2003a Bryde's whale 0.49 0.02 1.01 yes NE Pacific, west coast US 1991-1993 S Barlow 2003a Bryde's whale 0.49 0.05 1.07 no NW Atlantic, northern Gulf of Mexico (Oceanic Surveys) 1996-1997 S Davis et al. 2000 Bryde's whale 0.49 0.43 1.05 no NW Atlantic, northern Gulf of Mexico (GulfCet I EPA survey) 1996-1997 S Davis et al. 2000      Food web models and data for NW African ecosystem, L. Morissette et al.  18 Table 4 (cont.) Common Name Estimated density [animals / 1000 km2] Observed density [animals / 1000 km2] CV G(0) corrected Geographic area Survey years Survey type Source Fin whale 0.22 0.00 0.00 yes NOAA 91-96 California inshore 1991 A Forney et al. 1995 Fin whale 0.22 16.09 0.22 no NW Mediterranean 1992 S Forcada et al. 1995 Fin whale 0.22 1.97 0.35 yes NE Pacific, west coast US 1991-1993 S Barlow 2003a Fin whale 0.22 3.18 0.34 yes NE Pacific, west coast US 1996 S Barlow 2003a Fin whale 0.22 3.92 0.56 yes NE Pacific, west coast US 2001 S Barlow 2003a Fin whale 0.22 0.07 0.72 yes NE Pacific, Hawaiin waters 2002 S Barlow 2003b Fin whale 0.22 0.19 1.01 yes NE Pacific, California inshore 1991-1992 A Forney et al. 1995 Fin whale 0.22 0.07 1.15 no NW Atlantic, US east coast, south of Maryland 1998 S Mullin and Fulling 2003 Fin whale 0.22 1.85 0.48 no NW Atlantic, Virginia Capes 2002 S Garrison et al. 2003 Humpback whale 0.10 1.22 0.41 yes NE Pacific, California inshore 1991-1992 A Forney et al. 1995 Humpback whale 0.10 3.50 0.21 yes NE Pacific, California inshore 1991-1996 S Calambokidis and Barlow 2004 Humpback whale 0.10 0.03 0.37 yes NE Pacific, California offshore 1991-1996 S Calambokidis and Barlow 2004 Humpback whale 0.10 1.46 0.42 yes NE Pacific, California waters 1991 A Forney and Barlow 1993 Humpback whale 0.10 0.00 0.00 yes NE Pacific, Baja California 1993 S Calambokidis and Barlow 2004 Humpback whale 0.10 0.66 0.41 yes NE Pacific, west coast US 1991-1993 S Barlow 2003a Humpback whale 0.10 0.14 0.72 yes NE Pacific, west coast US 1996 S Calambokidis and Barlow 2004 Humpback whale 0.10 1.81 0.44 yes NE Pacific, west coast US 1996 S Barlow 2003a Humpback whale 0.10 0.89 0.49 yes NE Pacific, west coast US 2001 S Barlow 2003a Humpback whale 0.10 7.44 0.48 no SE Atlantic, Gabon waters (northern strata) 2002 A Rosenbaum et al. 2004 Humpback whale 0.10 24.34 0.31 no SE Atlantic, Gabon waters (southern strata) 2002 A Rosenbaum et al. 2004 Humpback whale 0.10 46.49 0.47 no SW Indian Ocean, Madagaskar (eastern block) 1994 S Best et al. 1996 Humpback whale 0.10 112.32 0.27 no SW Indian Ocean, Madagaskar (southern block) 1994 S Best et al. 1996 Humpback whale 0.10 67.47 0.15 no SW Indian Ocean, Mozambique 2003 S Findlay et al. 1994        Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  19 Table4 (cont.) Common Name Estimated density [animals / 1000 km2] Observed density [animals / 1000 km2] CV G(0) corrected Geographic area Survey years Survey type Source Minke whale 1.28 0.42 0.68 yes NE Pacific, California waters 1991 A Forney and Barlow 1993 Minke whale 1.28 0.28 0.62 yes NE Pacific, California inshore 1991-1992 A Forney et al. 1995 Minke whale 1.28 0.27 0.44 yes NE Pacific, west coast US 1991-1993 S Barlow 2003a Minke whale 1.28 0.93 0.51 yes NE Pacific, west coast US 1996 S Barlow 2003a Minke whale 1.28 0.86 0.77 yes NE Pacific, west coast US 2001 S Barlow 2003a Minke whale 1.28 0.03 1.29 no NW Atlantic, US east coast, south of Maryland 1998 S Mullin and Fulling 2003 Sei whale 0.10 0.06 0.53 yes NE Pacific, west coast US 1991 S Barlow 2003a Sei whale 0.10 0.05 0.79 yes NE Pacific, west coast US 1991 S Barlow 2003a Sei whale 0.10 0.10 0.73 yes NE Pacific, west coast US 1996 S Barlow 2003a Sei whale 0.10 0.03 1.01 yes NE Pacific, west coast US 2001 S Barlow 2003a Sei whale 0.10 0.03 1.06 yes NE Pacific, Hawaiin waters 2002 S Barlow 2003b Killer whale 0.11 0.44 0.37 no Eastern Tropical Pacific 1986-1990 S Wade and Gerrodette 1993 Killer whale 0.11 0.31 0.76 yes NE Pacific, California waters 1991 A Forney and Barlow 1993 Killer whale 0.11 0.25 0.69 yes NE Pacific, California inshore 1991-1992 A Forney et al. 1995 Killer whale 0.11 0.55 0.50 yes NE Pacific, west coast US 1991-1993 S Barlow 2003a Killer whale 0.11 0.74 0.61 yes NE Pacific, west coast US 1996 S Barlow 2003a Killer whale 0.11 0.58 0.73 yes NE Pacific, west coast US 2001 S Barlow 2003a Killer whale 0.11 0.14 0.98 yes NE Pacific, Hawaiin waters 2002 S Barlow 2006 Killer whale 0.11 0.69 0.42 no NW Atlantic, northern Gulf of Mexico (Oceanic Surveys) 1991-1994 S Hansen et al. 1995 Killer whale 0.11 0.79 0.48 no NW Atlantic, northern Gulf of Mexico (GulfCet I survey) 1991-1994 S Davis and Fargion 1996 Killer whale 0.11 0.17 1.01 no NW Atlantic, northern Gulf of Mexico (Oceanic Surveys) 1996-1997 S Davis et al. 2000 Killer whale 0.11 0.37 0.49 no NW Atlantic, northern Gulf of Mexico (SEFSC) 1996-2001 S Mullin and Fulling 2004  Food web models and data for NW African ecosystem, L. Morissette et al.  20 Table 4 (cont.) Common Name Estimated density [animals / 1000 km2] Observed density [animals / 1000 km2] CV G(0) corrected Geographic area Survey years Survey type Source Sperm whale 1.09 1.33 0.22 yes Eastern Tropical Pacific 1986-1990 A & S Wade and Gerrodette 1993 Sperm whale 1.09 0.92 0.38 no Eastern Tropical Pacific 1998 S Gerrodette and Forcada 2002 Sperm whale 1.09 1.24 0.60 no Eastern Tropical Pacific 1999 S Gerrodette and Forcada 2002 Sperm whale 1.09 0.19 0.73 no Eastern Tropical Pacific 2000 S Gerrodette and Forcada 2002 Sperm whale 1.09 3.36 0.81 yes Northeastern Tropical Pacific 1997 S Barlow and Taylor 2005 Sperm whale 1.09 4.10 0.36 no Northeastern Tropical Pacific 1997- 2000 S Barlow and Taylor 2005 Sperm whale 1.09 0.20 1.07 yes NE Pacific, California waters 1991 A Forney and Barlow 1993 Sperm whale 1.09 3.42 0.99 yes NE Pacific, California inshore 1991-1992 A Forney et al. 1995 Sperm whale 1.09 1.41 0.40 yes NE Pacific, west coast US 1991-1993 S Barlow 2003a Sperm whale 1.09 0.47 0.56 yes NE Pacific, west coast US 1996 S Barlow 2003a Sperm whale 1.09 1.90 0.59 yes NE Pacific, west coast US 2001 S Barlow 2003a Sperm whale 1.09 0.90 0.13 yes NE Pacific, Hawaiin waters 1993-1998 A Mobley et al. 2000 Sperm whale 1.09 2.79 0.81 yes NE Pacific, Hawaiin waters 2002 S Barlow 2006 Sperm whale 1.09 2.06 0.51 no NW Atlantic, US east coast, south of Maryland 1998 S Mullin and Fulling 2003 Sperm whale 1.09 1.14 0.85 no NW Atlantic, Virginia Capes 2002 S Garrison et al. 2003 Sperm whale 1.09 2.31 0.31 yes NW Atlantic, northern Gulf of Mexico (SEFSC) 1991-1994 S Hansen et al. 1995 Sperm whale 1.09 1.31 0.31 no NW Atlantic, northern Gulf of Mexico (Oceanic Surveys) 1991-1994 S Hansen et al. 1995 Sperm whale 1.09 1.74 0.30 no NW Atlantic, northern Gulf of Mexico (GulfCet I survey) 1991-1994 S Davis and Fargion 1996 Sperm whale 1.09 0.96 0.45 no NW Atlantic, northern Gulf of Mexico (Oceanic Surveys) 1996-1997 S Davis et al. 2000 Sperm whale 1.09 1.62 0.56 no NW Atlantic, northern Gulf of Mexico (GulfCet I survey) 1996-1997 S Davis et al. 2000 Sperm whale 1.09 0.85 0.57 no NW Atlantic, northern Gulf of Mexico (GulfCet I EPA survey) 1996-1997 S Davis et al. 2000 Sperm whale 1.09 3.80 0.23 no NW Atlantic, northern Gulf of Mexico (SEFSC) 1996-2001 S Mullin and Fulling 2004  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  21 1. Minke whales Minke whales (Balaenoptera acutorostrata) are the most abundant species of whales in our study area. According to quantitative estimations by Kaschner (2004), there is an average abundance of 4574 whales off the coast of Northwest Africa annually, translating to a density of 1.28 individuals per 1000 km2 and representing a total biomass of 30,050 tonnes, or an annual biomass density of 0.00844 t*km-2. The predicted density (individuals per 1000 km2) was in the same range as observed densities reported from dedicated marine mammals surveys conducted in other areas with similar habitats (Table 4). The annual mortality of minke whales was estimated by Evans (1998) to be around 0.09 – 0.10. We assumed this stayed constant over the study period, and also that following Allen (1971), this mortality equals the production to biomass ratio (P/B) needed for model’s construction. This was similar to the P/B values specifically used for Minke whales in another Ecopath model (Guénette et al. 2006). A P/B value of 0.099 yr-1 was used here. Unfortunately, there are few quantitative descriptions of diet for cetaceans in general, and Minke whales are no exception. Since no study has been conducted on minke whale diet in Northwest African countries, we used a set of six papers published in the literature about the diet of Minke whales in the North Atlantic (Table 5). The average diet was used as an input in our Ecopath model, while minimum and maximum values were used as ranges for calibration. Table 5. Available information on the diet composition of minke whales in the North Atlantic. The average diet was used in the Ecopath model for Northwest Africa. Prey groups (Ecopath) Northeast Atlantic (Nørdoy and Blix, 1992) Northeast Atlantic (Haug et al. 1996) North Sea (Olsen and Holst 2001) Norway (Lydersen et al. 1991) Norwegian waters (Haug et al. 1995) Norwegian waters (Olsen and Holst 2001) Norwegian waters (Smout and Lindstrom 2007) Eastern Norway (Sivertsen 2006) Avg. diet 14. Bathydemers al predators 0.053 0.232  0.003 0.078   0.067 0.054 18. Coastal demersals 0.025 0.134 0.896 0.081 0.132    0.159 19. Clupeids 0.501 0.358 0.011 0.916 0.382 1.000 0.039 0.303 0.439 20. Coastal pelagics  0.031 0.093  0.326  0.222 0.153 0.103 25. Zooplankton 0.421 0.245   0.082  0.739 0.477 0.245 Total 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000  2. Fin whales Fin whales (Balaenoptera physalus) are another important species of baleen whales in the area. Kaschner (2004) estimated their abundance to 780 individuals in the area of Northwest Africa, translating to a density of 0.22 individuals per 1000 km2, representing a total biomass of 43,372 tonnes, or an annual biomass density of 0.012 t*km-2. The predicted density (individuals per 1000 km2) was similar in terms of magnitude as observed densities reported from dedicated marine mammals surveys conducted in other areas with similar habitats (Table 4). The generally accepted natural mortality rate for adult fin whales ranges from 0.04 to 0.06 (Clark 1982; de la Mare 1985; Perry et al. 1999). Total mortality was assumed to be a bit higher, but no information was available about that. Consequently, we used a P/B value of 0.099 yr-1, following Heymans’ (2005) estimate for the same species in the Aleutian Islands and the Gulf of Alaska. No information was available on the diet of fin whales in Northwest Africa, so we used the study by Sigurjónsson and Víkingsson (1997) on the feeding of fin whales in Icelandic waters (Table 6). Fin whales were mostly feeding on zooplankton, but fish remains were also found in their stomach. Food web models and data for NW African ecosystem, L. Morissette et al.  22 Table 6. Diet composition of fin whales used for Ecopath model Northwest Africa. Prey groups (Ecopath) Diet (Sigurjónsson and Víkingsson 1997) 18. Coastal demersals 0.002 20. Coastal pelagics 0.014 25. Zooplankton 0.984 Total 1.000  3. Humpback whales Humpback whales (Megaptera novaeangliae) also occur in the waters of Northwest Africa. Quantitative estimations by Kaschner (2004) showed that there is an average abundance of 354 whales in our study area annually, representing a density of 0.10 individuals per 1000 km2, for a total biomass of 10,770 tonnes, or an annual biomass density of 0.0030 t*km-2. The predicted density (individuals per 1000 km2) was similar in terms of magnitude to the range of observed densities reported from dedicated marine mammals surveys conducted in other areas with similar habitats (Table 4). Following Heymans (2005), we used a P/B ratio of 0.099 yr-1 for humpback whales. This value was used in the Aleutian Islands and the Gulf of Alaska for the same species, and in absence of any other information on total mortality of humpback whales, we assumed that this ratio was similar in Northwest Africa. The only information available on the diet of humpback whales in Atlantic waters came from Mitchel (1973). However, since this study only listed proportions of “krill” and “fish” as prey in the diet, we used a more recent study by Witteveen et al. (2006) to distribute “fish” prey into more precise categories. Consequently, the diet of humpback whales used in the Northwest Africa model was mainly composed of zooplankton, coastal pelagic, and coastal demersal fish (Table 7). Table 7. Diet composition of humpback whales used for Ecopath model Northwest Africa. Prey groups (Ecopath) Diet (Sigurjónsson and Víkingsson 1997, after Mitchel 1973) 14. Bathydemersal predators 0.086 18. Coastal demersals 0.171 19. Clupeids 0.086 20. Coastal pelagics 0.275 25. Zooplankton 0.400 Total 1.000  4. Bryde’s whales The distribution of Bryde’s whale, Balaenoptera brydei, in Northwest Africa is scarcely documented. However, there were reported sightings of this species in the waters of Morocco and in the Caribbean by Kato (2002). From the quantitative estimates of Kaschner (2004), the annual abundance of this species in the region is approximately 1,731 individuals or a density of 0.49 individuals per 1000 km2, with a total biomass of 27, 961 tonnes or a biomass density of 0.00785 t*km-2. Predicted and observed densities reported from dedicated marine mammals surveys conducted in other areas with similar habitats were similar in terms of magnitude (Table 4). In the absence of mortality value or P/B ratio of this species in the region, an average P/B ratio of 0.099 yr-1 for baleen whales in Aleutian Islands, Alaska (Guénette et al., 1996) was used in the present model. The dietary information for the Bryde’s whale is lacking in the region. Thus, we used diet information of this species from Best (2001) in South Africa. It was documented that Bryde’s whale feed mainly on Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  23 zooplankton (60.3%), other coastal pelagics (28.8%), mesopelagic predators (5.0%), clupeids (3.2%) and bathydemersal predators (2.7%) (Table 8). Table 8. Diet composition of Bryde’s whales used for Ecopath model Northwest Africa. Prey groups (Ecopath) Diet (Best 2001) 13. Mesopelagic predators  0.050 14. Bathydemersal predators 0.027 19. Clupeids 0.032 20. Coastal pelagics 0.288 25. Zooplankton 0.603 Total 1.000  5. Sei whales Sei whales, Balaenoptera borealis, are known to inhabit the waters of Northwest Africa during their breeding season (Klinowska 1991). Population of this species in the region had been reduced because of commercial whaling in the early 1950s (Klinowska 1991; Horwood 2002). Based on the recent quantitative estimate of sei whales in Northwest Africa by Kaschner (2004), their annual total abundance is approximately 339 individuals , translating to a density of 0.10 individuals per 1000 km2, and a biomass of 5,697 tonnes or a biomass density of 0.0016 t*km-2. The predicted densities (individuals per 1000 km2) were similar in terms of magnitude as observed densities reported from dedicated marine mammals surveys conducted in other areas with similar habitats (Table 4). Minke whales (Balaenoptera acutorostrata) are the most abundant species of whales in our study area. According to quantitative estimations by Kaschner (2004), there is an average abundance of 4574 whales off the coast of Northwest Africa annually, translating to a density of 1.28 individuals per 1000 km2 and representing a total biomass of 30,050 tonnes, or an annual biomass density of 0.00844 t*km-2.  The P/B ratio of 0.020 yr-1 used for sei whales in the present model for Northwest Africa was obtained from the Alaska Gyre ecosystem model (Pauly et al. 1996). There is no existing dietary information of sei whales in the region. Hence, we used the diet information of this species from the North Atlantic (Sigurjónsson and Víkingsson 1997; Mitchell 1974). The dietary contents of sei whales were mainly composed of zooplankton (98.0%) and a few species of fish (Table 9). Table 9. Diet composition of sei whales used for Ecopath model Northwest Africa. Prey groups (Ecopath) Diet (Sigurjónsson and Víkingsson 1997) 14. Bathydemersal predators 0.007 18. Coastal demersals 0.007 20. Coastal pelagics 0.007 25. Zooplankton 0.980 Total 1.000  6. Blue whales The species of baleen whales included in this group is the blue whale (Balaenoptera musculus). This group inhabits the Cape Verde Islands during breeding seasons. According to Kaschner’s estimates (2004), this group represents a total abundance of 96 individuals annually, or 0.03 individuals per 1000 km2, representing a biomass 9,831 tonnes, or a biomass density 0.00276 t*km-2 in the region. The predicted density (individuals per 1000 km2) was similar in terms of magnitude as observed densities reported from dedicated marine mammals surveys conducted in other areas with similar habitats and showed in Table 4. Food web models and data for NW African ecosystem, L. Morissette et al.  24 The P/B ratio of 0.04 yr-1 used for the baleen whale in the present model is the average value taken from many Ecopath models of this group (Guénette and Christensen 2005; Okey 2004; Pauly et al. 1996). This value ranges from 0.01 to 0.05 yr-1, depending on the source and baleen whales species included in these models. In absence of dietary information of baleen whale species (blue whale) in Northwest Africa, we incorporated diet results of blue whales from Icelandic waters (Sigurjónsson and Víkingsson 1997). Blue whales consume mainly zooplankton (Sigurjónsson and Víkingsson 1997) (Table 10). This was also mentioned by several experts (e.g. Hjort and Ruud 1929; Klinowska 1991; Sears 2002; Hewitt and Lipsky 2002) that describe the ecology and distribution of blue whales. Table 10. Available information on the diet composition of blue whales used for Ecopath model in the Northwest African ecosystem. Prey groups (Ecopath) Diet (Sigurjónsson and Víkingsson 1997) 25. Zooplankton 1.000 Total 1.000  7. Sperm whales Sperm whale (Physeter macrocephalus) abundance was calculated by Kaschner (2004) to be around 3878 whales off the coast of Northwest Africa annually, representing a density of 1.09 individuals per 1000 km2, for a total biomass of 71,865 tonnes, or an annual biomass density of 0.020 t*km-2. This estimated density (in individuals per 1000 km2) is similar in terms of magnitude as observed densities reported from dedicated marine mammals surveys conducted in other areas with similar habitats (Table 4). The annual mortality of sperm whales was estimated by Perry et al. (1999) to be around 0.05. In the absence of any other information, we assumed this to be equivalent to the P/B for that species (following Allen 1971). Thus, the P/B value of our model was 0.05 yr-1. Most publications on sperm whales diet indicate that they feed primarily on cephalopods (Kawakami 1980; Clarke et al. 1993; González et al. 1994; Roberts 2003). However, a study by Best (1999) also includes fish prey in the diet. In any case, no information was available specifically for the Northwest Africa region, and we thus used an average diet based on all information available in North Atlantic waters (Table 11). The average diet was used as an input in our Ecopath model, while minimum and maximum values were used as ranges for calibration. Table 11. Available information on the diet composition of sperm whales in the North Atlantic. The average diet was used in the Ecopath model for Northwest Africa. Prey groups (Ecopath) West coast of South Africa (Best 1999) Vigo, Spain (Kawakam i 1980) Madera, Spain (Kawakam i 1980) Azores (Clarke et al. 1993) South Crête, Greece (Roberts 2003) Northwest Spanish Atlantic coast (González et al. 1994) Avg. diet 12. Large pelagics 0.014      0.002 13. Mesopelagic predators 0.014      0.002 14. Bathydemersal predators 0.014      0.002 21. Cephalopods 0.876 1.000 1.000 1.000 1.000 1.000 0.979 22. Crustaceans 0.041      0.007 23. Benthos 0.041      0.007 Total 1.000 1.000 1.000 1.000 1.000 1.000 1.000  8. Killer whales Killer whales (Orcinus orca) are the top predators in the Northwest African ecosystem, occupying the highest trophic level of the food web. Kaschner (2004) estimated their abundance as 409 individuals in Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  25 the area of Northwest Africa annually, based on a estimated density of 0.11 individuals per 1000 km2, for a total biomass of 933 tonnes, or an annual biomass density of 0.00026 t*km-2. Predicted density (individuals per 1000 km2) was similar in terms of magnitude as observed densities reported from dedicated marine mammals surveys conducted in other areas with similar habitats (Table 4). The P/B value we used for the model was taken from Sidi and Guénette (2004) who calculated a ratio of 0.02 yr-1 based on a calculation by Trites and Heise (1996). Diet composition of this group was also based on Sidi and Guénette (2004) for Mauritania (after a study by Paul et al. 1998), who estimated that killer whales mainly feed on dolphins, coastal demersals, and cephalopods (Table 12). Table 12. Diet composition of killer whales used for Ecopath model Northwest Africa. Prey groups (Ecopath) Diet (Sidi and Guénette 2004) 10. Dolphins 0.384 12. Large pelagics 0.010 14. Bathydemersal predators 0.010 18. Coastal demersals 0.379 19. Clupeids 0.048 20. Coastal pelagics 0.038 21. Cephalopods 0.131 Total 1.000  9. Beaked whales Three species of beaked whales were included in that group: Cuvier’s beaked whale (Ziphius caviostris), Blainville’s beaked whale (Mesoplodon densirostris), and Gervais’ beaked whale (M. europaeus). Kaschner (2004) estimated a total annual abundance of 693 beaked whales, representing a biomass 446 tonnes, or a biomass density 0.0001 t*km-2. The P/B ratio was taken from Heymans (2005) who used a value of 0.036 for beaked whales species in their model of the Aleutian Islands. Information on the diet of beaked whales species were available for all three species included in our model. Here again, the diet was calculated as a weighted average of these three diets, based on the fact that Cuvier’s beaked whales consume 66% of the food in our study area, while Blainville’s beaked whales consume 24%, and Gervais’ beaked whales consume 10% (Kaschner 2004). The resulting diet composition is mainly made up of cephalopods and bathydemersal predators (Table 13). Here again, the average diet was used as an input in our Ecopath model, while minimum and maximum values were used as ranges for calibration. Table 13. Available information on the diet composition of beaked whales in the North Atlantic. A weighted average diet (based on the % of food consumption by each species) was used in the Ecopath model for Northwest Africa. Prey groups (Ecopath) Cuvier’s beaked whale Canary Islands (Santos et al. 2007) Blainville’s beaked whale Canary Islands (Santos et al. 2007) Gervais’ beaked whales Canary Islands (Santos et al. 2007) Weighted average diet 13. Mesopelagic predators  0.002 0.291 0.029 14. Bathydemersal predators  0.847  0.207 21. Cephalopods 1.000 0.152 0.710 0.763 Total 1.000 1.000 1.000 1.000  Food web models and data for NW African ecosystem, L. Morissette et al.  26 10. Small cetaceans A total of 17 species of dolphins and other small toothed whales are included in that group (Table 3). Most species are currently observed in the Northwest Africa area, but some are rarely observed or only in stranding events (Sidi and Guénette 2004). The database compiled by Kaschner (2004) allowed us to estimate a total annual abundance of 180,735 individuals, for a total biomass of 19,921 tonnes, or a density of 0.0055 yr-1. The original model by Samb and Mendy (2004) included a specific trophic group for dolphins, so we used the P/B value initially presented there for our model. Consequently, the value of 0.0047 yr-1 is used in this broader model and assumed to represent all species of small toothed cetaceans and dolphins of the Northwest African area. Diet information about dolphins was also taken directly from the initial model by Samb and Mendy (2004). The diet was based on Northridge (1984) and mainly composed of coastal fish as well as zooplankton (Table 14). Table 14. Diet composition of dolphins used for Ecopath model Northwest Africa. Prey groups (Ecopath) Diet (Samb and Mendy 2004) 18. Coastal demersals 0.080 19. Clupeids 0.291 20. Coastal pelagics 0.098 21. Cephalopods 0.193 25. Zooplankton 0.339 Total 1.000  Food consumption by marine mammals Kaschner (2004) developed a basic food consumption model based on Trites et al. (1997). This model was used to generate the biomasses and consumption (Q/B) values needed for each Ecopath group. Annual food consumption was calculated as: ! Q i = 365* N i,s W i,s R i,s s " where the annual food consumption Q of species i was assumed to be 365 times the daily food consumption. Daily food consumption is calculated based on the number of individuals N of the sex s of a species i, and a weight-specific daily ration R consumed by an individual with a species- and sex-specific mean body mass. Abundances and sex ratios were taken directly from the Kaschner (2004) database. Mean species and sex- specific body mass was taken from Trites and Pauly (1998). For all cetaceans, except baleen whales, we used the empirical model developed by Innes et al. (1987) to estimate food consumption of cetaceans that was later modified by Trites et al. (1997) to account for the difference between consumption for growth and for maintenance and then applied to all marine mammal species. Food intake of specific species per day was calculated using: 8.0 s.isi, W*0.1R = where R is the daily food intake of an individual of sex s belonging to species i and W  is the mean body weight of that individual, in kilograms. For all baleen whales daily food ration was estimated based on a model by Armstrong and Siegfried (1991) for food consumption of minke whales in the Antarctic. These authors suggested a modification to the Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  27 empirical model of Innes et al. (1986) equation for baleen whales to account for larger body sizes and seasonal variation in food intake. This approach was later used to estimate food consumption of whales around Iceland (Sigurjónsson and Víkingsson 1997) and represents one of the methods used by Tamura (2003) to estimate global food intake of cetaceans. This feeding rate is calculated as: ! R i,s = 0.42*W i,s 0.67 Annual food consumption for each species of marine mammals was then divided by the biomass estimates, in order to get the final consumption to biomass (Q/B) ratios used in the Ecopath model (Table 15). Table 15. Consumption estimates for each marine mammal group of the Ecopath model of Northwest Africa. Ecopath groups Annual food consumption (tonnes*km-2) Annual biomass (tonnes*km-2) Q/B 1. Minke whales 0.071 0.00844 8.4212 2. Fin whales 0.051 0.01220 4.1608 3. Humpback whales 0.014 0.00302 5.0781 4. Bryde’s whales 0.049 0.00785 6.2600 5. Sei whales 0.010 0.00160 6.1776 6. Blue whales 0.009 0.00276 3.3978 7. Sperm whales 0.102 0.02020 5.0251 8. Killer whales 0.002 0.00026 7.7634 9. Beaked whales 0.001 0.00013 9.9234 10. Dolphins 0.306 0.00559 13.7390 Average 0.045 - 6.995  11. Seabirds Seabirds of Northwest Africa are migrating species and are mostly present in the study area from November until March (Samb and Mendy 2004). Very little information is available for seabirds species in that area, and input values in the original model came from southern Benguela (Crawford et al., 1991). The input values for that group are thus a total biomass of 0.118 t*km-2, a P/B of 0.12 year-1, and a Q/B of 118 year-1. Diet composition was mainly composed of zooplankton, pelagic fish, and other small fish (Jarre-Teichmann et al. 1998). 12. Large pelagics This group is mainly composed of large tunas that are important target of “surface” fisheries, and bycatch by longlines. The most important species of this group are yellowfin (Thunnus albacares), skipjack (Katsuwonus pelamis), and bigeye (Thunnus obesus) tunas. These species are large and migrating species that are mainly caught by a global fishery that happens outside Northwest African countries’ EEZ. Samb and Mendy (2004) estimated a biomass density of 2.54 t*km-2 for the Senegambian ecosystem. This value was assumed to represent our study area of Northwest Africa. The biomass value for large pelagic species could range from 0.427 to 5.346 t*km-2 in other models for the Northwest African region. Thus, this represents an intermediate value. The average annual catch of large pelagic fish in the 1990s was 0.0078 t*km-2 for local fleets, and 0.0075 t*km-2 for foreign fleets, for an average total catch of 0.0153 t*km-2 for the 1990s in Northwest Africa. Natural mortality for tropical tunas was estimated to range between 0.6 to 0.8 yr-1 (Samb and Mendy 2004). Total mortality (Z) of yellowfin tuna was estimated to 1.6 yr-1 (ICCAT 1999a,b), following Allen (1971). This represents a higher value compared to other P/B values used for large pelagics in similar Food web models and data for NW African ecosystem, L. Morissette et al.  28 Ecopath models in Northwest Africa, which range from 0.300 yr-1 (Diallo et al. 2004) to 1.908 yr-1 (Stanford et al. 2001). An average value of 1.016 yr-1 was used here. Q/B ratio of large pelagic species range from 3.560 yr-1 (Sidi and Guénette 2004 for Mauritania) to 34.490 yr-1 (Stobberup et al. 2004 for Cape Verde) in Northwest Africa. For the present model, we used an average Q/B of 11.698 yr-1. Diet information was taken from Cayré et al. (1988) who estimated that large pelagic tunas feed mainly on fish, crustaceans, and molluscs.  13. Mesopelagic predators This group was not originally included in the Senegambian Ecopath model of Samb and Mendy (2004), but was added because fisheries data indicate that mesopelagic predators are commercially important in the study area. The key species of that group for Northwest Africa are blue whiting (Micromesistius poutassou) and black scabbardfish (Aphanopus carbo). The biomass of mesopelagic predators was taken from similar species group of the Guinea-Bissau model (Amorim et al. 2004) and was set to 0.735 t*km-2yr-1. The P/B ratios used in different Ecopath models of the Northwest Africa area were ranging from 3.990 yr- 1 in Mauritania (Sidi and Guénette 2004) to 5.105 yr-1 in Morocco (Stanford et al. 2001). An average value of 4.362 yr-1 was used for the present model. The Q/B ratios available for Northwest Africa were ranging from 4.440 yr-1 in Mauritania (Sidi and Guénette 2004) to 37.506 yr-1 in Morocco (Stanford et al. 2001) An average value of 31.609 yr-1 was used for the model covering the greater area of Northwest Africa. The diet of mesopelagic predators also comes from the Guinea-Bissau model and is assumed to be representative of the larger ecosystem of Northwest Africa. According to Amorim et al. (2004), mesopelagic predators mainly fed on large pelagic fish, clupeids, and cephalopods.  14. Bathydemersal predators Here again, the bathydemersal predators group was not initially included in the Senegambian Ecopath model of Samb and Mendy (2004) because this model was only covering the continental shelf, and most of the bathydemersal predators are found in deeper waters. Since our study area covers deeper waters, we added a trophic group to cover these species. The key species of that group for Northwest Africa are Senegalese hake (Merluccius senegalensis), European hake (Merluccius merluccius) and Black-bellied angler (Lophius budegassa). The biomass of bathydemersal predators was taken from a similar species group of the Cape Verde model (Stobberup et al. 2004) and was set to 0.255 t*km-2yr-1. Over the whole study area, this value could range from 0.255 to 0.729 t*km-2yr-1. The P/B ratios used in different Ecopath models of the Northwest Africa area were ranging from 0.384 yr-1 (Stobberup et al. 2004) to 2.710 yr-1 (Sidi and Guénette 2004). The lowest value of 0.384 yr-1, taken from the Cape Verde model (Stobberup et al. 2004), was assumed to represent the Northwest Africa region, since it was calculated for the same trophic group (bathydemersals) in both models. The Q/B ratio used in this model also comes from the Cape Verde model (Stobberup et al. 2004) and was set to 3.844 yr-1. This was the lowest value of the range of Q/B for Northwest Africa models. Maximum Q/B observed in the area was 17.43 yr-1 in Mauritania (Sidi and Guénette 2004). Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  29 The diet of bathydemersal predators also comes from the Cape Verde model and is assumed to be representative of the larger ecosystem of Northwest Africa. According to Stobberup et al. (2004), bathydemersal predators mainly feed on large crustaceans, coastal demersal fish, zooplankton, and benthos.  15. Sharks Most shark species of the area belong to the Carcharhinidae family. The key species of that group is the blue shark, Prionace glauca. Species belonging to the Squalidae family (dogfish sharks) as well as mustelus (smooth-hounds) species are also important. Surveys by the Centre de Recherche Océanographique de Dakar-Thiaroye (CRODT) evaluated the biomass to be around 5,600 tonnes for the Senegambian area, equivalent to a biomass density of 0.290 t*km-2. According to mission reports from Norwegian vessel Fultjort Nansen and their 2007 survey for the Northwest African area (unpublished data), this biomass decreased to 3470 tonnes in FAO area 34 (or approximately 0.126 t*km-2). The biomass value could vary from 0.144 (Sidi and Guénette 2004) to 0.729 (Stanford et al. 2001) in our study area. The average annual catches were estimated by SAUP to 0.0011 t*km-2 for local fleets and 0.0004 t*km-2 for foreign fleets for the 1990s period, for a total catch of 0.0015 t*km-2*yr-1 for sharks. These values represent a fishing mortality of 0.0074 yr-1 over the study area. The biology of sharks remains generally unknown in the study area. P/B (or total mortality) was estimated from growth parameters of Carcharhinus limbatus presented in Bransletter (1987). The author estimated the L∞ to be 176 cm FL (or 210 cm TL), and a K of 0.27 yr-1. Using the equation of Pauly (1980), these growth parameters and an average temperature of 22˚C, a natural mortality of 0.28 yr-1 was estimated by Samb and Mendy (2004). This natural mortality added to fishing mortality give a Z or P/B of 0.905 yr-1 which was used in the model. This value could vary from 0.300 yr-1 (Diallo et al. 2004) to 1.338 yr-1 (Stanford et al. 2001) in other published models of our study area. The Q/B of sharks in Northwest Africa could range from 3.000 yr-1 (Mendy 2004) up to 11.477 yr-1 (Stanford et al. 2001) for sharks. An average value of 7.497 yr-1 was used here and assumed to represent the entire area of Northwest Africa. Diet of sharks was based on Samb and Mendy (2004), and was mainly composed of zooplankton and coastal clupeid fish.  16. Rays Rays were grouped separately from sharks because of their particular ecology. Indeed, since these species live on the bottom of the ocean, their feeding behaviour is different than that of sharks. The key species of this group is the common guitarfish (Rhinobatos rhinobatos). For the 1990s, Samb and Mendy (2004) estimated an average annual biomass of 0.112 t*km-2, which is what we used in the model. However, in 2007, this biomass increased to 4340 tonnes (or approximately 0.157 t*km-2) in the area, according to the mission reports from Norwegian vessel Fultjort Nansen survey off the coast of Northwest Africa in 2007 (I.L. Bamy, Centre National des Sciences Halieutiques de Boussoura, unpublished data). Catch data came from the Sea Around Us database (R. Watson, Sea Around Us Project, personal communication), and resulted in a total of 0.0018 t*km-2yr-1 of rays caught in the 1990s for our study area. This catch was almost equally divided between local fleets (0.0010 t*km-2yr-1) and foreign fleets (0.0008 t*km-2yr-1). Based on what was available for the common guitarfish, Samb and Mendy (2004) estimated a P/B of 1.000 year-1 based on maximal length of 100 cm (after Schneider 1990) and an average temperature of 22°C, using Pauly’s (1980) equation to estimate natural mortality. Total mortality (Z) was then obtained Food web models and data for NW African ecosystem, L. Morissette et al.  30 by adding fishing mortality to the natural mortality. This value was assumed to equal the production rate, based on Allen (1971). The range of possible P/B used in other models of the Northwest Africa region was from 0.423 yr-1 (Stobberup et al. 2004) to 1.000 yr-1 (Samb and Mendy 2004). A median of 0.755 yr-1 was used here as a starting point. Q/B was averaged from similar models (De la Cruz-Aguero 1993, Opitz 1993, Paula e Silva et al. 1993) and set to 6.000 year-1. This value can range from 3.912 yr-1 in Guinea-Bissau (Amorim et al. 2004) to 6.3 yr-1 in Mauritania (Sidi and Guénette 2004) for similar Ecopath models. Diet composition was mainly composed of benthos, but also includes a small proportion of coastal demersal species.  17. Coastal tunas This group includes a dozen species, of which the most important are little tunny (Euthynnus alletteratus), Atlantic bonito (Sarda sarda), West African Spanish mackerel (Scomberomorus tritor), and Frigate tuna (Auxis thazard). These species are present all year round in the Northwest African waters, and are predominantly fished by artisanal fisheries. The biomass of coastal tunas was taken directly from the estimate of Samb and Mendy (2004) for the Senegambian ecosystem and was set to 2.89 t*km-2yr-1. The average annual catches were estimated by SAUP to 0.0013 t*km-2 for local fleets and 0.0017 t*km-2 for foreign fleets for the 1990s period, for a total catch of 0.0030 t*km-2*yr-1 for coastal tunas. Growth parameters have been estimated for the little tunny by Pauly (1978, 1979), Diouf (1980), and Pauly and Munro (1984). Samb and Mendy (2004) used these parameters to estimate the mortality of this group, based on Pauly’s (1980) equation. The M was calculated to 0.34 yr-1 based on growth parameters and water temperature of 22˚C. Assuming that fishing mortality was equal to their exploitation rate of 0.5 yr-1, a Z value of 0.800 was estimated for coastal tunas in the area. This value was assumed to represent the average coastal tunas in the ecosystem and was then used as an input in the Northwest African model. Other P/B values in Northwest African models could range from 0.642 yr-1 (Stanford et al. 2001) to 0.987 yr-1 (Stobberup et al. 2004). The Q/B ratio used in this model also comes directly from the Senegambian model (Samb and Mendy 2004) and was set to 9.500 yr-1. This was close to the highest value of the range of Q/B for Northwest Africa models (with a maximum Q/B of 9.872 yr-1 in Cape Verde [Stobberup et al. 2004]). Minimum Q/B observed in the area was 3.774 yr-1 in Morocco (Stanford et al. 2001). Diet of coastal tunas was based on Samb and Mendy (2004), who cite a study by Postel (1955) describing the diet of coastal tunas to be mainly composed of small coastal pelagics, crustaceans, and cephalopods.  18. Coastal demersals demersal species differ a lot depending on the habitat type and depth. They include a wide range of species that are subject to an important level of exploitation (Samb and Mendy 2004). Two types of coastal demersals are included in that group. Fish living on soft sediments in warmer estuarine waters with high productivity belong to the sciaenidae (Pseudotolithus spp.), polynemidae (Galeoides spp.), carangidae (Scyris spp.) and cynoglossidae (Cynoglossus spp.) families. Other fish live on diverse sediment types, varying from rocks to mud, and from depths from 30 to 100 meters, in colder waters. The latter species belong to the sparidae (Dentex spp.), serranidae (Epinephelus spp.), sparidae (Pagellus spp. and Sparus spp.) and mullidae (Pseudupaeneus spp.) families. Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  31 The coastal demersal fish are highly exploited in the Northwest African area. Consequently, our data collection allowed us to realize that most information available on fisheries or biomass surveys are for these species. According to Samb and Mendy (2004), most of the region’s coastal demersal species are overexploited. Based on a synthesis of eight trawl surveys in the senegambian area (Caverivière and Thiam 1992), the exploitable biomass declined substantially from 1986 to 1991, going from 173,000 to 81,000 tonnes. The biomass of coastal demersal fish in the Senegambian waters was estimated to 4.696 t*km-2 by Samb and Mendy (2004). Assuming this density represents the whole Northwest African region, this value is in the lower range of possible biomass used in other Ecopath models published in the same area, ranging from 2.634 t*km-2 (Sidi and Guénette 2004) to 27.267 t*km-2 (Stanford et al. 2001). The biomass of this species is also known to be decreasing. According to a survey of the Norwegian vessel Fultjort Nansen off the coast of Northwest Africa in 2007 (Mission reports, I.L. Bamy, Centre National des Sciences Halieutiques de Boussoura, unpublished data), the biomass of the major demersal groups in Northwest Africa was 49,860 tonnes (or 1,765 t*km-2). The P/B ratio used in our model was calculated from an estimation of natural mortality of 0.4 yr-1 (Franqueville 1983) for red Pandora (Pagellus bellottii). Using that M, Samb and Mendy (2004) estimated a natural mortality (M) of 0.5 yr-1, based on Pauly (1980) equation and with a L∞ of 37 cm (LT) and a K of 0.24 yr-1, for an average temperature of 22˚C. When this M is added to a calculated fishing mortality of 0.7 yr-1 (Samb and Mendy 2004), this results in a P/B of 1.200 yr-1. This represents the lowest value of the range of possible values found for Northwest Africa models. The highest part of the range was 13.940 yr-1 in Guinea (Diallo et al. 2004). The Q/B ratio used in this model comes from an estimate by Samb and Mendy (2004) using an equation from Palomares and Pauly (1998), resulting in a value of 6.000 yr-1. This represents the lowest value from the range of possible Q/B for coastal demersal species in Northwest Africa. The highest Q/B value found for this species group was 57.279 yr-1 in Guinea-Bissau (Amorim et al. 2004) An extreme Q/B value of 107.636 yr-1 (Stobberup et al. 2004) was not considered in the range. The diet of coastal demersal fish is as diverse as the species composing that group. The main prey of this group include crustaceans, zooplankton, pelagic fish, cephalopods, benthos, and detritus (Fischer et al. 1981, Franqueville 1983, Sidibé 2003) According to Samb and Mendy (2004), coastal demersal fish feed mainly on benthic producers and plankton.  19. Clupeids Clupeids represent with other coastal pelagics the most important marine resource in terms of landings in the area. Depending on the years, coastal pelagics could reach 70% of declared catches in the Senegambian area (Samb 1997). The most important clupeid species of the area are round and madeiran sardinellas (Sardinella aurita and Sardinella maderensis, respectively), which are considered as the key species for the clupeids group. Bonga shad (Ethmalosa fimbriata) is also an important species of coastal and estuarine zones of our study area. The biomass of sardinellas in the area results from the N/O Dr Fridtjof Nansen survey that performed acoustic survey campaigns each November from 1995 to 1999 in the EEZ of Senegal and Gambia (Saetersdal et al. 1995; Toresen 1996, 1997, 1998; Toresen and Kolding 1999; Samb and Pauly 2000). Average biomass for the round sardinella was 237,000 tonnes, and it was 373,000 tonnes for the madeiran sardinella. In the 2007 survey, the biomass of these species was 248,000 and 328,000 tonnes, respectively (Mission reports, I.L. Bamy, Centre National des Sciences Halieutiques de Boussoura, unpublished data). Using the 1990s biomass, we ended up with a total annual biomass density of 22.130 t*km-2 for Northwest Africa. Catch data were collected from the SAUP database and represent an average of 0.1931 t*km-2 for local fleets and 0.1423 t*km-2 for foreign fleets, totaling an average 0.3354 t*km-2 of clupeids caught annually in the area in the 1990s. Food web models and data for NW African ecosystem, L. Morissette et al.  32 Natural mortality for sardinellas was estimated by many authors in the area, notably Camarena-Luhrs (1986), Samb (1988), and Fréon (1988). Samb and Mendy (2004) used a natural mortality of 0.96 yr-1 for round sardinella and 0.5 yr-1 for madeiran sardinella. Added to fishing mortality, the authors ended up with a P/B of 1.54 yr-1 for round sardinella and 0.72 yr-1 for madeiran sardinella. Since this group was aggregated in the present model, a weighted average P/B of 1.33 yr-1 was used here. In other published models from the same study area, this value could range from 1.100 yr-1 (Stanford et al. 2001) to 3.100 yr-1 (Diallo et al. 2004). The Q/B ratio for sardinellas was also averaged from Samb and Mendy (2004), and represents 16.347 yr-1. The authors used Q/B ratios of 20.200 yr-1 and 13.900 yr-1 for round sardinella and madeiran sardinella, respectively. In other models published for our study area, these values could range from 9.000 yr-1 in Gambia (Mendy 2004) to 28.389 yr-1 in Guinea-Bissau (Amorim et al. 2004). Diet of sardinellas was mainly composed of zooplankton and phytoplankton (Nieland 1982; Medina- Gaertner 1985).  20. Other coastal pelagics The key species of that group are ray-finned fish species such as the chub mackerel (Scomber japonicus), horsemackerels (Trachurus spp.), and largehead hairtail (Trichiurus lepturus). Samb and Mendy (2004) mention that for the Senegambian area, the commercially important species in terms of catches are the cunene horse mackerel (Trachurus trecae) and the false scad (Decapterus rhonchus). Other coastal pelagic species are also caught, including bigeye grunt (Brachydeuterus auritus), Atlantic bumper (Chloroscombrus chrysurus), and sompat grunt (Pomadasys jubelini). An average biomass of 13.116 t*km-2 was used by Samb and Mendy (2004), based on the survey of the N/O Dr Fridtjof Nansen during the early 2000s. The total biomass was estimated to be around 362,000 tonnes for carangidae and other coastal pelagic species in the Senegambian area. In 2007, this biomass was down to 224,000 for FAO area 34 (Mission reports, I.L. Bamy, Centre National des Sciences Halieutiques de Boussoura, unpublished data). The density estimated by Samb and Mendy (2004) in the Senegambian area was assumed to be representative of our study area in the late 1980s. The average total catch of coastal pelagic species in the 1990s was estimated to reach 502,153 tonnes per year, for a density of 0.1410 t*km-2 in our study area. From this amount, 0.0924 t*km-2 were caught by foreign fleets, and 0.0487 t*km-2 by local fleets. Samb and Mendy (2004) used a total mortality (equal to P/B) of 1.100 yr-1, based on a natural mortality of 0.5 yr-1 for cunene horse mackerel, estimated by Maxim (1995). This P/B was an intermediate value and was assumed to be representative of the whole Northwest African area, which could range from 0.600 yr-1 (Mendy 2004) to 4.380 yr-1 (Sidi and Guénette 2004), according to other published Ecopath models in the area. The consumption rate of coastal pelagic species used in our model was 10.635 yr-1, based on Samb and Mendy (2004). However, this could range from 4.440 yr-1 (Amorim et al. 2004) to 41.050 yr-1 (Sidi and Guénette 2004) from other models published in the study area. Diet of this trophic grouping was mainly composed of zooplankton, and also included some phytoplankton (Samb and Mendy 2004).  21. Cephalopods Cephalopods are important species for fisheries in Northwest Africa, especially in Morocco and Mauritania (Samb and Mendy 2004). Octopus (Octopus spp.) and cuttlefish (Sepia spp.) are the dominant species of cephalopods caught in the study area, but other species such as European common Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  33 squid (Alloteuthis subulata), short-finned squid (Illex coindetii), European squid (Loligo vulgaris), European flying squid (Todarodes sagittatus) are also included in this group. Depending on the years we see important fluctuations in the biomass of squids in the study area. The average biomass in the Senegambian area was estimated to 30,000 tonnes for the 1990s. This resulted in a biomass density of 1.087 t*km-2, assumed to be representative of the whole study area. However, in 2007, a survey by the Norwegian vessel Fultjort Nansen off the coast of Northwest Africa (Mission reports, I.L. Bamy, Centre National des Sciences Halieutiques de Boussoura, unpublished data), estimated the biomass to be down to 3,630 tonnes (approximately 0.131 t*km-2) for FAO area 34 (Northwest Africa). As mentioned by many authors, cephalopods are an important commercial species for all Northwest African countries. Total catches in our study area averaged 193,953 tonnes per year in the 1990s, representing a density of 0.0545 t*km-2. The P/B of cephalopods is relatively unknown. In the study area, this value could range from 1.900 yr-1 (Samb and Mendy 2004) to 4.700 yr-1 (Diallo et al. 2004). An average value of 3.300 yr-1 was used to represent our study area. Similarly little information was available for the Q/B of cephalopods in Northwest Africa. Other published Ecopath models used values ranging from 11.700 yr-1 in Senegambian and Moroccan ecosystems (Samb and Mendy 2004; Stanford et al. 2001) to 23.400 yr-1 (Sidi and Guénette 2004). An average of 16.735 yr-1, based on three models of Northwest Africa (Samb and Mendy 2004; Amorim et al. 2004; Stanford et al. 2001; Sidi and Guénette 2004), was used in our model. Diet of cephalopods was mainly composed of zooplankton, macrobenthos, and also included some fish species (Samb and Mendy 2004, after Diata et al. 2001).  22. Crustaceans This group was not originally included in the model of Samb and Mendy (2004), but was added because an important fishery occurs on shrimp (Parapenaeus longirostris and Penaeus spp.). Biomass was obtained from similar crustacean species of the Cape Verde model (Stobberup et al. 2004), and was set to 13.048 t*km-2*yr-1. Based on the SAUP database, we used an average total catch of 27,159 tonnes of crustaceans per year in the 1990s for Northwest Africa. This represents a density of 0.0076 t*km-2 for our model. Crustaceans are mainly caught by local fleets (0.0062 t*km-2), but also by foreign fleets (0.0014 t*km-2). P/B ratio could range from 2.500 yr-1 (Diallo et al. 2004) to 15.190 yr-1 (Amorim et al. 2004). The average P/B ratio available among four models in the whole Northwest Africa area was 6.443 yr-1. This was used as the input value in our model. Q/B ratios are ranging from 10.000 yr-1 (Stobberup et al. 2004) to 54.840 yr-1 (Amorim et al. 2004) in different models available from Northwest Africa. An average value of 31.963 yr-1 based on five models of the same area, was used as an input here. The diet of large crustaceans also comes from Cape Verde and is assumed to be representative of the larger ecosystem of Northwest Africa. According to Stobberup et al. (2004), crustaceans mainly feed on benthos and detritus.  23. Benthos Macrobenthos is an important source of food for demersal fish, but knowledge about this group is very limited (Samb and Mendy 2004). In the original model, this group was divided between macrobenthos Food web models and data for NW African ecosystem, L. Morissette et al.  34 and meiobenthos, and the different parameters were based on a model from Jarre-Teichmann (1998). In the present model, macro and meiobenthos are grouped, and the different input data are aggregated and weighted by their biomass. Total biomass density for benthos was estimated to 112.600 t*km-2. The P/B could range from 1.687 yr-1 (Samb and Mendy 2004) to 7.000 yr-1 (Sidi and Guénette 2004). An extreme value of 108.300 yr-1 (Amorim et al. 2004) was not included in the range here. An average value of 4.33 yr-1 was used in our model. The range of possible Q/B values provided from other models in Northwest Africa was from 13.481 yr-1 (Samb and Mendy 2004) to 48.670 (Sidi and Guénette 2004). Here again, an extreme value of 243.100 yr-1 (Amorim et al. 2004) was not included in the average. In the present model, we used an average value of 32.730 yr-1, based on three models from Northwest Africa. Based on information on macrobenthos and meiobenthos groups in the model by Samb and Mendy (2004), our aggregated ‘Benthos’ group was assumed to have an average diet mainly composed of detritus (84%), but also benthic producers and other benthic species (11%).  24. Benthic producers Since data about benthic production is very sparse, the original model of Samb and Mendy (2004) obtained information on benthic production from another ecosystem model (southern Benguela, by Jarre- Teichmann et al. 1998). We decided to leave that group unmodified, and thus used a biomass value of 10.500 t*km-2 and a P/B value of 15.000 yr-1 for the current model.  25. Zooplankton Research on secondary production, mostly done in waters surrounding Cape-Verde, suggest a strong relationship between the abundance of zooplankton and the strength of the upwelling. In this area, copepods compose the major part of the biomass, which totals 20.636 t*km-2 (based on Touré 1983). Based on Samb and Mendy (2004), we used a P/B of 58.356 yr-1 and a Q/B of 274.805 yr-1. The latter value may seem high, but in fact, it falls within the range of what is usually seen in the literature for zooplankton’s consumption. Other published Ecopath models of the northwestern African region present P/B values ranging from 28.000 yr-1 in Mauritania (Sidi and Guénette 2004) to 78.356 yr-1 in Guinea (Diallo et al. 2004), and Q/B values ranging from 106.330 yr-1 in Mauritania (Sidi and Guénette 2004) to 280.000 yr-1 in Cape Verde (Stobberup et al. 2004).  26. Phytoplankton Primary production was calculated by Samb and Mendy (2004) as an average of different estimates. Primary production can be very important in the Northwest African area because it’s highly influenced by the upwelling that makes the area very rich in oxygen and have a high productivity (Voiturier and Herbland 1982). In the area, the maximum production is observed at the surface, where maximal values of 70 to 80 mg*m-2 has been recorded (Voiturier and Herbland 1982). We used an average biomass value of 82.000 t*km-2 and a P/B of 138.189 yr-1, based on Samb and Mendy (2004).  27. Detritus Information is very sparse about the biomass of detritus on the coast of Northwest Africa. As a result, an arbitrary total biomass of 10 t*km-2 was initially used by Samb and Mendy (2004). The lack of information on detritus is very common for other Ecopath models. Most of the time, a very generalized value is Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  35 assumed (Morissette et al. 2003; Morissette 2005). Detritus is assumed to include benthic and pelagic detritus that fall on the bottom of the ocean, as well as benthic bacteria involved in the microbial loop (Morissette et al. 2003).  BALANCING THE MODEL The unbalanced model for Northwest Africa is shown in Table 16. In order to obtain a balanced solution, different levels of verification have been made. First, it was important to make sure that gross efficiency (GE), which is the ratio of production to consumption (P/Q) was always within the 0.1 – 0.3 range. According to Christensen and Pauly (1992), GE ranges between 10 and 30%, with the exception of top predators, e.g., marine mammals and seabirds, which can have lower GE (between 0.1 and 1%), and small, fast growing fish larvae or nauplii or bacteria, which can have higher GE (between 25% and 50%) (Christensen and Pauly 1992). Table 16. Input data for the Ecopath model of Northwest Africa. Unbalanced values are shown in bold. Ecopath group Biomass (t*km-2) P/B (year-1) Q/B (year-1) EE GE 1. Minke whales 0.00844 0.0990 8.4212 0.0000 0.0118 2. Fin whales 0.0122 0.0990 4.1608 0.0000 0.0238 3. Humpback whales 0.00302 0.0990 5.0781 0.0000 0.0195 4. Brydes whales 0.00785 0.0990 6.2600 0.0000 0.0158 5. Sei whales 0.00160 0.0200 6.1776 0.0000 0.0032 6. Sperm whales 0.0202 0.0500 5.0251 0.0000 0.0100 7. Killer whales 0.000262 0.0200 7.7634 0.0000 0.0026 8. Baleen whales 0.00276 0.0400 3.3978 0.0000 0.0118 9. Beaked whales 0.000125 0.0360 9.9234 0.0000 0.0036 10. Dolphins 0.00559 0.0470 13.7390 2.9729 0.0034 11. Seabirds 0.1180 0.1200 118.0000 0.0000 0.0010 12. Large pelagics 2.5400 1.0160 11.6980 0.9474 0.0869 13. Mesopelagic predators 0.7350 4.3620 31.6090 0.1170 0.1380 14. Bathydeersal predators 0.2550 0.3840 3.8440 7.6971 0.0999 15. Sharks 0.2900 0.9050 7.4970 0.0106 0.1207 16. Rays 0.1120 0.7550 6.0000 0.0216 0.1258 17. Coastal tunas 2.8900 0.8000 9.5000 0.0026 0.0842 18. Coastal demersal 4.6960 1.2000 6.0000 0.3142 0.2000 19. Clupeids 22.1300 1.3300 16.3466  1.3062 0.0814 20. Other coastal pelagics 13.1160 1.1000 10.6350 0.9500 0.1034 21. Cephalopods 1.0870 3.3000 16.7350 2.3499 0.1972 22. Crustaceans 13.0480 6.4430 31.9630 0.0053 0.2016 23. Benthos 112.6000 4.3340 32.7300 0.6620 0.1324 24. Benthic producers 10.5000 107.3600 - 0.2634 - 25. Zooplankton 20.6360 58.3560 274.8050 0.4015 0.2124 26. Phytoplankton 82.0140 138.1890 - 0.5109 - 27. Detritus 10.0000 - - 0.3672 -  The modifications to the original model needed to reach a balanced solution are listed below: Food web models and data for NW African ecosystem, L. Morissette et al.  36 1. The biomass for dolphins originally came from an estimate by Northridge (1984), and had a higher value (0.039 t*km-2). In our modified version of the model, biomass was calculated from K. Kaschner’s database, and had a value of 0.0059 t*km-2. However, because dolphins are also an important prey of killer whales in this ecosystem, the lower biomass of dolphins created an imbalance (too much predation for the available biomass and production), and an ecotrophic efficiency (EE) higher than 1. To reach a balanced solution, an intermediate value of 0.02225 t*km-2 (the average of the minimal and maximal value) was used for dolphins. 2. The group representing large pelagics was not balanced in our model. Consequently, we used the highest possible P/B and Q/B values for this group, equivalent to 1.908 yr-1 and 11.60 yr-1, respectively. This resulted in a balanced solution and allowed us to keep our initial biomass of 2.54 t*km-2. 3. Bathydemersal predators is a group that was added to the Senegalo-Gambian ecosystem to represent an important fishery for the coast of Northwest Africa. Since this group was not in the initial model from Samb and Mendy (2004), we first assumed that the biomass of bathydemersal predators would be the same in terms of density (0.255 t*km-2) than what is presented for Cape Verde by Stobberup et al. (2004). However, this seems to be inadequate to reach a balance solution, so we let the model estimate the biomass needed to be at equilibrium with the rest of the foodweb (using a EE of 0.95). Consequently, we used the highest P/B value available in our range (2.71, from Sidi and Guénette 2004), and let the model estimate the Q/B using the assumption that P/Q ratio is 30% (Christensen and Pauly 1992). The estimated biomass for bathydemersal predators is 0.2928  t*km-2 and the Q/B is 9.033 yr-1,  which are still within the range of possible values for similar species living on the coast of Northwest Africa. 4. Clupeids were clearly overestimated in our model, even if their EE was close to balance in our model, with a value of 1.28. In order to reach a balanced solution, we reduced the proportion of clupeids in the predators’ diet, used a maximal P/B value of 3.100 yr-1 from the Ecopath model of Guinea (Diallo et al. 2004). This was sufficient to balance the model. 5. In our unbalanced model, there were not enough ‘other coastal pelagic’ fish to reach a balanced solution. We thus used an average P/B value of 2.61 yr-1 and this was sufficient to reach a balanced scenario. 6. Many species feed on cephalopods in our ecosystem, and consequently the input parameters used as a starting point were not sufficient to cope with the consumption by all predators of these species in our model. Thus, we used a maximum P/B of 4.700 yr-1 from the Guinean model (Diallo et al. 2004). We also used a biomass of 2.4092 t*km-2 to be consistent with the available time series of biomass for this species in the area. This biomass represents an intermediate within the range of possible values computed for all models for Northwest Africa (ranging from 1.038 t*km-2 in Mauritania [Sidi and Guénette 2004] to 4.507 t*km-2 in Guinea-Bissau [Amorim et al. 2004]), and is assumed to be representative of the area. 7. Benthic producers had a low P/B of 15 yr-1 in our ecosystem, and this created an imbalance in the initial parameterization of our model. However, we know that the P/B of benthic producers is higher in other models such as Guinea-Bissau (Amorim et al. 2004) where they used a P/B of 199 yr-1, and could even reach 213 yr-1 in Cape Verde (Stobberup et al. 2004). We thus increased the production rate of benthic producers to an intermediate value of 107.362 yr-1. This produced a balanced solution for that group. 8. All these adjustments created an imbalance on other coastal pelagics (EE = 1.352), so we used an average P/B of 2.612 yr-1 (within the range of possible values for Northwest Africa) to reach a balanced solution for that group. 9. Finally, coastal tunas had a GE lower than 10%, so we used the highest possible P/B (0.820 yr-1 from Sidi and Guénette 2004), combined with the lowest possible Q/B (3.774 yr-1 from Stanford et al. 2001) to re-balance the model. The resulting GE was 22% which is within the range of what Christensen and Pauly (1992) suggested. Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  37 The final balanced model for Northwest Africa is given in Table 17. A diet matrix showing the proportion of each prey in all predators’ diets is provided in Table 18.  Table 17. Balanced Ecopath model of Northwest Africa. Estimated parameters are shown in bold. Ecopath group Trophic level Biomass (t*km-2) P/B (year-1) Q/B (year-1) EE GE 1. Minke whales 3.54 0.00844 0.0990 8.4212 0.0000 0.0118 2. Fin whales 3.01 0.0122 0.0990 4.1608 0.0000 0.0238 3. Humpback whales 3.44 0.00302 0.0990 5.0781 0.0000 0.0195 4. Brydes whales 3.39 0.00785 0.0990 6.2600 0.0000 0.0158 5. Sei whales 3.02 0.00160 0.0200 6.1776 0.0000 0.0032 6. Sperm whales 4.12 0.0202 0.0500 5.0251 0.0000 0.0100 7. Killer whales 3.94 0.000262 0.0200 7.7634 0.0000 0.0026 8. Baleen whales 3.00 0.00276 0.0400 3.3978 0.0000 0.0118 9. Beaked whales 4.21 0.000125 0.0360 9.9234 0.0000 0.0036 10. Dolphins 3.54 0.0225 0.0470 13.7390 0.7386 0.0034 11. Seabirds 3.78 0.1180 0.1200 118.0000 0.0000 0.0010 12. Large pelagics 3.31 2.5400 1.9080 11.6980 0.5045 0.1631 13. Mesopelagic predators 3.95 0.7350 4.3620 31.6090 0.1170 0.1380 14. Bathydeersal predators 3.40 0.2928 2.7100 9.0333 0.9500 0.3000 15. Sharks 3.43 0.2900 1.3380 7.4970 0.0116 0.1785 16. Rays 2.65 0.1120 1.0000 6.0000 0.0318 0.1667 17. Coastal tunas 3.35 2.8900 0.8200 3.7740 0.0026 0.2173 18. Coastal demersal 2.28 4.6960 13.9400 46.4667 0.0492 0.3000 19. Clupeids 2.77 22.1300 3.1000 16.3466 0.5217 0.1896 20. Other coastal pelagics 2.80 13.1160 2.6120 10.6350 0.4112 0.2456 21. Cephalopods 3.13 2.4092 4.7000 16.7350 0.7445 0.2808 22. Crustaceans 2.67 13.0480 6.4430 31.9630 0.0140 0.2016 23. Benthos 2.02 112.6000 4.3340 32.7300 0.7222 0.1324 24. Benthic producers 1.00 10.5000 107.3600 - 0.3266 - 25. Zooplankton 2.00 20.6360 58.3560 274.8050 0.4266 0.2124 26. Phytoplankton 1.00 82.0140 138.1890 - 0.5151 - 27. Detritus 1.00 10.0000 - - 0.3695 -     Food web models and data for NW African ecosystem, L. Morissette et al.  38 Table 18. Diet matrix for the Ecopath model of Northwest Africa. Prey / Predators 1 2 3 4 5 6 7 8 9 10 11 1. Minke whales 2. Fin whales 3. Humpback whales 4. Brydes whales 5. Sei whales 6. Sperm whales 7. Killer whales 8. Baleen whales 9. Beaked whales 10. Dolphins       0.384 11. Seabirds 12. Large pelagics      0.002 0.010 13. Mesopelagic predators    0.050  0.002   0.029 14. Bathydeersal predators 0.005  0.009 0.027 0.001 0.002 0.010  0.207 15. Sharks 16. Rays 17. Coastal tunas 18. Coastal demersal 0.016 0.000 0.017  0.001  0.379   0.080 19. Clupeids 0.044  0.009 0.032   0.048   0.291 0.667 20. Other coastal pelagics 0.010 0.001 0.025 0.288 0.001  0.038   0.098 0.333 21. Cephalopods      0.979 0.131  0.764 0.193 22. Crustaceans      0.007 23. Benthos      0.007 24. Benthic producers 25. Zooplankton 0.025 0.098 0.040 0.603 0.098   0.100  0.339 26. Phytoplankton 27. Detritus Import 0.900 0.900 0.901  0.900   0.900 Total 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  39 Table 18 (cont.). Diet matrix for the Ecopath model of Northwest Africa. Prey / Predators 12 13 14 15 16 17 18 19 20 21 22 23 1. Minke whales 2. Fin whales 3. Humpback whales 4. Brydes whales 5. Sei whales 6. Sperm whales 7. Killer whales 8. Baleen whales 9. Beaked whales 10. Dolphins 11. Seabirds 12. Large pelagics  0.100  0.050 13. Mesopelagic predators  0.016 14. Bathydeersal predators  0.032 15. Sharks 0.000     0.000 0.000 16. Rays 0.000     0.000 0.000 17. Coastal tunas 18. Coastal demersal  0.016 0.367 0.100 0.140     0.037 19. Clupeids 0.158 0.580  0.275  0.308 0.002   0.090 20. Other coastal pelagics 0.158   0.150  0.103 0.005   0.050 21. Cephalopods 0.053 0.256    0.026 0.002 22. Crustaceans   0.443 23. Benthos   0.055  0.460  0.125   0.250 0.560 0.022 24. Benthic producers     0.400  0.375    0.200 0.055 25. Zooplankton 0.632  0.135 0.425  0.564 0.140 0.774 0.800 0.573 0.101 26. Phytoplankton       0.255 0.226 0.200  0.005 27. Detritus       0.096    0.134 0.923 Import Total 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 Food web models and data for NW African ecosystem, L. Morissette et al.  40 The changes in the general trophic structure from the unbalanced model to balanced models are presented below. For both scenarios the system is dominated by coastal fish (coastal tunas, clupeids, and other coastal pelagics), and the ‘offshore species’ group is the one that changed the most in terms of biomass.  This is due to the increase of biomass of two offshore species groups that were initially unbalanced: large pelagics and bathydemersal predators, for which we let the model estimate the biomass needed to cope for all the mortalities, using an ecotrophic efficiency of 95%. In both cases, the biomass of these groups almost doubled, but was still within the range of possible values for similar species in other models of Northwest Africa.      TIME SERIES DATA Biomass Few time series data on biomass were available for Northwest African waters. Martial Laurens (2005) published his PhD thesis on stock assessments of marine resources of western Africa. In his thesis, time series of biomass of different coastal demersal fish are presented. However, all the species covered by his Figure 4. Comparison of the biomass structure (not including benthic invertebrates & plankton) of the unbalanced vs balanced Ecopath models. Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  41 analysis fall into our ‘coastal demersals’ trophic group: white grouper (Epinephelus aenus), red Pandora (Pagellus bellottii), bluespotted seabream (Pagrus caeruleostictus), West African goatfish (Pseudupeneus prayensis), and lesser African threadfin (Galeoides decadactylus). We summarized these data into an aggregated time series representing coastal demersals (Table 19). Gascuel et al. (2007) estimated the biomass of 24 different demersal taxa, selected for their importance in the survey catches and/or the fishery in Mauritania. This was presented as year-to-year fluctuations from 1982 to 2006. Different Ecopath trophic groups were covered by this study: bathydemersal predators (Merluccius sp.), sharks (Mustelus mustelus), Rays (Raja miraletus, R. straeleni), and cephalopods (Octopus vulgaris, Sepia officinalis). The resulting time series, in density (t*km-2), is presented in table 19. A report of the FAO (2002) presents the outcomes of the Working Group on the Assessment of Small Pelagic Fish off Northwest Africa, after their workshop in Banjul, Republic of the Gambia, 5-12 April 2002. In that document, indices of biomass were calculated for Sardinella aurita and S. maderensis from acoustic surveys of the R/V by Dr. Fridtjof Nansen. The surveyed area covered the coastal shelf and immediately adjacent waters and ranges from Safi, in Morocco, to Ghana (from about 32°N to 5°N) a distance of some 2,500 nmi. This is similar to what our study area covers. After converting the biomass of these acoustic surveys into densities (t*km-2), we ended up with time series for our “clupeid” group for the Ecopath model (Table 19). Another report by the International Commission for the Conservation of Atlantic Tunas (ICCAT 1999c) estimated the biomass trends for skipjack tuna (Katsuwonus pelamis), which is one of the key species of our “large pelagics” group for the model. As an alternative to the traditional production models used by the Standing Committee on Research and Statistics (SCRS), a new model was presented by Dr. Maury during an inter-sessional stock assessment session of Atlantic skipjack in Funchal, Madeira, 28 June to 2 July 1999. This original non-equilibrium production model was developed for skipjack stocks, and although various runs of this experimental production model were made, only the results from a total Atlantic stock model were presented and discussed. It was considered that the results obtained from this model are probably dominated by the eastern Atlantic fisheries, which are consistent with our study area of Northwest Africa. The resulting time series of biomass densities for the large pelagics are presented in table 19. Additionally, a report by Chassot et al. (2006) presented a global dynamic model developed in a Bayesian context to evaluate the stock of common octopus (Octopus vulgaris) in Mauritania from 1971 to 2005. Biomasses estimated by the model include uncertainty linked to natural variability, and error in observation to give a realistic representation of the dynamics of common octopus stocks in the area. According to these data, the biomass declined by more than 80% between 1971 and 2005. Since the study by Gascuel et al. (2007) presented more cephalopods species, and because the trend of declining biomass is similar in both studies (see Figure 5), we decided not to use the time series of Chassot et al. (2006) a priori. However, these data were considered as alternative values for model validation.  Food web models and data for NW African ecosystem, L. Morissette et al.  42   Finally, Gascuel and Laurans (2003) published a stock status report for 11 stocks of Guinea, Senegal and Cape Verde. A working group met at Mindelo in October 2001 for the ‘Système d’information et d'analyse des pêches’ (SIAP) project, and focused on two categories of analysis: stock assessment using the global approach, and estimation of time series of abundance with general linear models. The pink lobster (Palinurus charlestoni) stock of Cape Verde was one of the species presented in that report, and is a key species in our “crustaceans” group for the Ecopath model. The resulting time series of biomass density are presented in Table 19. Figure 5. Biomass trends for cephalopods species off the coast of Mauritania. Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  43 Table 19. Time series of biomass for different trophic groups of our Ecopath model of Northwest Africa. Yea r 10  Large pelagics (t*km-2) 12 Bathy- demersal predators (t*km-2) 13   Sharks (t*km-2) 14   Rays (t*km-2) 16  Coastal demersals (t*km-2) 17   Clupeids (t*km-2) 19   Cephalopod s (t*km-2) 20   Crustaceans (t*km-2) 1986 0.0066 0.0910 0.6143 0.4567   2.4092 0.0022 1987 0.0068 0.0548 0.3103 0.4330 2.7609  1.7203 0.0026 1988 0.0068 0.0374 0.1114 0.7788 2.4289  2.0763 0.0016 1989 0.0067 0.0318 0.2590 0.6085 2.4811  1.3455 0.0017 1990 0.0068 0.0262 0.4066 0.4382 2.7366  0.6146 0.0011 1991 0.0068 0.0817 0.3367 0.5338 2.6307  1.0995 0.0004 1992 0.0062 0.0901 0.2102 0.8500 2.5328  1.6619 0.0007 1993 0.0065 0.0568 0.1833 0.1719 2.2504  1.1425 0.0006 1994 0.0061 0.0638 0.1693 0.1201 2.0164  1.2481 0.0006 1995 0.0056 0.0514 0.1843 0.1588 1.8904 2.8000 1.1685 0.0004 1996 0.0056 0.0750 0.2723 0.1558 1.7628 2.5190 0.7975 0.0005 1997 0.0034 0.0431 0.0954 0.1394 1.7329 1.4595 0.6968 0.0006 1998 0.0019 0.0646 0.0759 0.1112 1.5284 1.4595 0.9087 1999 0.0034 0.0240 0.1963 0.1450 0.5353 2.8973 0.4978 200 0  0.1082 0.1449 0.1223  2.1514 0.5194 2001  0.0836 0.1963 0.1614  2.5730 0.5689 2002  0.1455 0.1395 0.0924   0.6433 2003  0.0615 0.2464 0.0959   0.7969 2004  0.0754 0.0868 0.0804   0.6611 2005  0.0669 0.1257 0.0610   0.3718 2006  0.0195 0.3119 0.0092   0.3269  Effort For clupeids and other coastal pelagics (groups 19 and 20 in our model), we used a report from the FAO Working Group on the Assessment of Small Pelagic Fish off Northwest Africa (FAO 2003). This report provides an exhaustive analysis and trends in the basic data (landings, catch, effort, length distribution and age distribution) collected by each country, for sardine, sardinellas, horse mackerels and mackerel. The assessment of the stocks was based on a variety of methods, including analysis of long-term trends in fishery data (landings, effort, CPUE, etc.). Fernandez et al. (2005) monitored and analyzed the catches that were landed from the Spanish deepwater trawl fishery for black hake off Mauritania. This is a highly specialized fishery, with two species of black hake (Merluccius senegalensis and M. polli) constituting between 77–99% of total landings, which have annually averaged 9 300 tons over the past two decades.  Landings data from the fishery off Mauritania were collected between 1984 and 2001, by fishing expedition and vessel. The captains’ log was useful for analyzing the fishing effort by fishing expedition as measured in fishing days. This information was available for the period 1992 – 2001, and is assumed to represent the effort on bathydemersal predators (group #14) in Ecosim. To have a complete time series of effort, we completed it with catch/biomass values for 1986 – 1991. Effort data was available for ten different species of tuna from the Standing Committee on Research and Statistics (SCRS) of the International Commission for the Conservation of Atlantic Tunas (ICCAT – http://www.iccat.int1t2ce.asp). In Northwest Africa, tuna fisheries continue to occupy an important place in the fishing sector, in particular, at the socioeconomic level, due to the important volume of investments made, the large number of direct and indirect employment generated, and the diversity of the fishing Food web models and data for NW African ecosystem, L. Morissette et al.  44 methods, including artisanal fisheries, small boats, and industrial activities (ICCAT 2006). Countries from our study area that were covered by this database are Cape Verde, Ivory Coast, Ghana, Guinea, Morocco. The fishing effort on all tuna species provided a time series from 1986 to 1999 over the study area. The catch and effort time series for sardina in the different fleets and zones of Northwest Africa were available from FAO CECAF Scientific Committee (2002). When all zones considered, the fishing effort on sardine show a decline since the early 1990s, then an increase from 1997 to 2000. This report also mentions that the effort has been declining since the early 1980s. Consequently, we completed our time series to cover 1986-2001 with a F=C/B. For all other groups where we didn’t have fishing mortalities or effort time series available, the biomass and catch time series described above were used to estimate fishing mortality (C/B) for the years that biomass estimates were available, and projected for the years that they were not. When no complete time series of biomass data were available, we forced the model to fit the catch time series. This technique has been used in other Ecosim models when data was unavailable (see Heymans 2005 for an example).  Table 19. Time series of effort (standardized values based on days at sea data) for different trophic groups of our Ecopath model of Northwest Africa. Year 12  Deep pelagics (t*km-2) 14 Bathy- demersal predators (t*km-2) 15   Sharks (t*km-2) 16   Rays (t*km-2) 18  Coastal demersals (t*km-2) 19   Clupeids (t*km-2) 21   Cephalopods (t*km-2) 22   Crustaceans (t*km-2) 1986 1.0000 1.0000 1.0000 1.0000    1.0000 1987 1.0000 0.7868 1.0000 1.0000 1.0000   1.0000 1988 0.8819 1.0479 2.9096 0.9768 0.9919   1.0594 1989 0.8523 1.4457 1.2169 0.8870 0.9192   1.3051 1990 0.8361 1.6170 0.7086 2.2608 0.8864 1.0000 1.0000 1.4911 1991 0.7101 0.5380 0.5991 1.9476 1.0408 0.9926 1.0301 1.7772 1992 0.7913 0.5248 1.1402 1.5703 1.0453 0.8331 0.9612 1.3629 1993 0.7088 0.9000 1.1459 7.2792 1.0570 0.8662 0.9014 1.6574 1994 0.6439 0.8667 1.3779 13.1076 1.1969 0.9834 1.0475 1.8317 1995 0.8889 0.8500 1.3249 9.7551 1.1714 1.0793 1.1863 2.1599 1996 0.7435 0.6833 1.5928 10.2987 1.3728 0.8935 1.0571 2.5034 1997 1.5253 0.6333 7.6822 20.6519 1.4956 0.6578 1.2995 2.0445 1998 2.5972 0.4500 5.2338 18.5378 1.7000 0.8982 0.9706 2.6327 1999 1.3069 0.4000 2.3347 12.9123 4.6958 1.0193 1.0506 3.2340 2000 1.3069 0.4833 3.2508 16.6169 2.1598 1.4191 1.3185 2.8202 2001 1.3069 0.6833 2.7467 11.7156 2.3240 0.9675 0.8262 3.2831 2002 1.3069 0.1937 3.9135 16.0606 2.5566 5.8382 1.1528 1.5637 2003 1.3069 0.5977 1.7999 14.4493 2.6934 6.1333 1.1528 1.4884 2004 1.3069 0.4951 5.8154 15.0962 2.9124 6.5126 1.1528 1.2261 2005 1.3069 0.3961 11.5865 18.9650 3.1108 6.8288 1.1528 1.1965 2006 1.3069 0.2690 12.6787 19.9667 3.3069 7.1871 1.1528 1.1491  UNCERTAINTY ANALYSES Given the high level of uncertainty in data, parameterization and model structure (Plagányi and Butterworth 2004; Plaganyi 2007; Essington 2007), we conducted several levels of uncertainty analyses. Effort was focussed on appropriate data collection to assist in shedding light as to the most appropriate choice of model form to represent feeding behaviour. Three levels of uncertainty analyses were performed here. First, a simple sensitivity routine included in Ecopath was used to explore the effects of uncertainty Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  45 on the model results. A second uncertainty analysis was performed using Ecoranger, a resampling routine based on input probability distributions. Finally, the robustness of our models’ structure was tested with Ecosim by comparing predicted biomasses with time series of observed data.  Sensitivity analysis A sensitivity routine is included in Ecopath to allow users to explore the effects of uncertainty on the model results. The method is quite simple, and consists of plotting relative output changes against relative changes in the inputs. The routine varies all basic input parameters (biomass [B], production to biomass ratio [P/B], consumption to biomass ratio [Q/B], ecotrophic efficiency [EE]) in steps from -50% to +50% for each trophic group of the model, and then checks what effect each of these steps has for each of the input parameters on all of the “missing” basic parameters for each group in the system (Christensen et al. 2000). The output is then given as the proportion of the difference between the estimated and original parameter to the original parameter, and converted to a percentage (Christensen et al. 2000). This method only re-estimates the parameters for which no data was available, and that were left to be estimated by the model, using the mass-balance constraints. We conducted a sensitivity analysis for biomass, P/B, Q/B and EE input parameters.  Our results suggest that the sensitivity of these estimated parameters to a change in input values is relatively low (Appendix 1).  A 50% change in any of the input parameters of any trophic group generated an overall response of about 35% in the estimated parameters of other groups.  Most of the changes in biomass would have a greater effect on the EE of other trophic groups (a 50% change in biomass generates an average 20% change in the EE of other trophic groups). In average, a 50% change in input biomass of any trophic group generated a response of 30% in the estimated parameters of other groups. For P/B ratio, a change in input values is more likely to affect EE parameters (with an average 50% of change after a 50% change in P/B). Overall, a 50% change in the P/B inputs of any group would generate a response of less than 48% in the estimated parameters of other groups. Similarly, for Q/B input values, a 50% change of this value for any group would generate a response of less than 25% in the estimated parameters of other trophic groups. The highest impact of a 50% change in Q/B is seen in estimating EE (with an average response of 28% difference). Finally, a change in EE input values generated an average 38% change in the estimated parameters. The most important effect of such change was seen on estimated biomass parameters, with an average 40% change after the EE is modified by ± 50%. Overall, our sensitivity analysis suggests that potential errors in model results are approximately linearly related to potential error in model parameters, etc.  This result is consistent with those of Essington (2007).  This underscores the importance of enhancing the quality of data included in our model.  ‘Ecoranger’ analysis To account for the inherent uncertainty of input parameters, a resampling routine called Ecoranger is included  in the EwE software and accepts input probability distributions for the biomasses, consumption and production rates, ecotrophic efficiencies, catch rates, and diet compositions. Ecoranger then draws random input variables using the range of possible values for each parameter, and the resulting model is then evaluated (based on least sum of squared residuals and physiological and mass-balance constraints) (Christensen et al. 2005). Starting with the initial model and these setups, 10,000 models were run by Ecoranger, until 200 model runs passed the selection criteria, and the best fitting model for the coast of Northwest Africa was used for further analysis. Food web models and data for NW African ecosystem, L. Morissette et al.  46  Fitting the model to time series data The Ecosim model behaviour is based on a ‘foraging arena’ theory (Walters and Martell, 2004), which assumes that predator and prey behaviours cause partitioning of prey populations, which are either available or unavailable to predators. There is continuous change between these two stages for any given potential prey, whether it is hiding from predation in some refuge, or it is out to feed. This availability of prey to predators is called ‘vulnerability’ in Ecosim. Mackinson et al. (2003) demonstrated the importance of setting the vulnerabilities to fit model predictions to time-series data, as Ecosim predictions are very sensitive to this parameter. Using default values for v has strong implications for assumptions about species abundance relative to their carrying capacity (V. Christensen, Fisheries Centre, UBC, personal communication). Basically, it assumes that each group can at most increase the predation mortality they impose on their prey with a factor of 2.0 (the default v value). A lower value implies a donor driven density-dependant interaction. On the other hand, a higher value involves a predator driven density-independent interaction, in which predation mortality is proportional to the product of prey and predator abundance (i.e., Lotka-Volterra). This implies a high flux rate for prey species in and out of vulnerable biomass pools (Ainsworth 2006). Vulnerabilities were thus adjusted based on the specific ecology of each species or trophic groups (if their behaviour, niche, or diet make them more or less vulnerable to predators). Using the few time series of biomass available for the trophic groups in our model, we compared Ecosim’s projections with observed data, and adjusted v’s and other input parameters (within their range of uncertainty) until we obtained a model configuration that allowed us to reproduce as much as possible the trends in biomass. Using credible models that can reproduce observed historical response to disturbances such as fishing is a useful approach to validating our model in light of the highly uncertain data included in the model. Fitting time series data to model predictions therefore enhances our confidence about the possible impact of removing marine mammals in the ecosystem (Morissette 2007).  Simulating the removal of great whales in the ecosystem A hunting pattern was chosen which generated an important increase in the mortality on the marine mammals, in order to drive their populations close to extinction. Vasconcellos et al. (1997) showed that for fish species, a 5-fold increase in anthropogenic predation leads to serious depletion in a group. Also, such an extreme scenario is routinely applied to many fish populations and often associated with stock collapse (Patterson, 1992). For marine mammals, Morissette (2007) proposed that the same kind of increase in anthropogenic mortality is needed to simulate a crash in marine mammal biomass.  Thus, we employed a similar approach for our analysis.  A 100 years simulation was performed, and the biomass trends before and after the removal of whales were compared.  DISCUSSION A new dataset built for Northwest Africa All the data collected and presented in this report were used to build an Ecopath model assessing the interactions between marine mammals and fisheries in Northwest African waters. Data presented here represent the updated version of an initial dataset presented to local experts during a workshop held in Dakar, Sénégal, in May 2008 (“Whale and Fish interactions: Are great whales a threat to fisheries”, see http://www.lenfestocean.org/whales_fisheries.html). The initial model presented in Dakar was  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  47 Additional data that became available after this workshop were included in the model, and thus greatly improved the predictive ability of our model simulations. We received a great amount of feedback from local experts, fisherman, and stakeholders about our model development and methodology. We also identified additional data on biomass, diet, and fishing effort for many species included in the model. As a consequence, we think that we increased our accuracy to represent key ecological processes in Northwest African waters.  Strengths and weaknesses of these modelling efforts The synthesis of available ecosystem information is allowing to have a whole-system view using parameters that are basic to understanding populations and the ecosystem (Okey and Pauly 1999). However, models are not a perfect representation of the reality. The uncertainties remaining in the understanding of the ecosystem structure and function may occur for different reasons. In some case, it is just because no data exist on the key species. In other cases, aggregation of species within one ecological box is inappropriate. Finally unknown mechanisms can also occur in the ecosystem (Morissette et al. 2003). For example, in Ecopath models, the only mechanism used to represent interactions between species is direct consumption. It ignores the fact that consumers often do more than skim production off their prey; consumers can shift diet composition to species with lower productivity and alter the P/B ratio of the group (Ruesink 1998), or have behaviours that indirectly affect other species in the system. The structure of the model provides an overall view of the ecosystem and reveals the uncertainties that could be examined in future studies. Consequently, one of the most important questions that can be asked of Ecopath models is: in which portion of the food web are the dynamics most uncertain? (Ruesink 1998). In our case, addressing the uncertainties linked to cetaceans and commercially important fish species was very important. On the other hand, a common problem in ecosystem modelling is that less information is available for the lower trophic levels (Moreau et al. 1993; Walline et al. 1993; Lin et al. 1999). These recurrent gaps generally force modellers to rely heavily on the literature and arbitrary assumptions to construct the models (Moreau et al. 1993), and the Northwest African region was no exception to that. This emphasizes the need for an increased research effort into the biomass, production, consumption, and diet of the various species of the ecosystem, not only the larger fish. Finally, in order to have reliable prediction from the model’s simulations, it is very important ot have local time series data on biomass, catch, and fishing effort. Unfortunately, such time series are very scarce in the Northwest African region. Even if our different uncertainty analyses suggested that our major findings are unlikely to change if we change data inputs, local research efforts aiming to fill the gaps in our knowledge of data would be useful. In that sense, that report represents a great tool to identify these gaps.  ACKNOWLEDGEMENTS Authors would like to thank all the participants to the Dakar workshop for providing inputs and comments on the methods and data used for the construction of this model. Special thanks are also given to Martial Laurans, Didier Gascuel, and Jean-Claude Brêthes for providing data and support for constructing this preliminary version of the model.  REFERENCES Aguilar A. and Borrell, A. 2007. Open-boat whaling on the straits of Gibraltar ground and adjacent waters. Marine Mammal Science 23: 322-342 Ainsworth, C.H. 2006. Strategic marine ecosystem restoration in Northern British Columbia. PhD thesis, University of British Columbia, Canada. Allen, K.R. 1971. Relation between production and biomass. Journal of Fisheries Research Board of Canada 28: 1573-1581. Amorim, P., Duarte, G., Guerra, M. Morato, T., and Stobberup, K.A. 2004. Preliminary Ecopath model of the Guinea-Bissau continental shelf ecosystem (NW Africa), pp. 95-112. In Palomares, M.L.D., and Pauly, D. (eds.) West African marine ecosystems: models and fisheries impacts. Fisheries Centre Research Reports 12(7). Armstrong, A.J., Siegfried, W.R. 1991. Consumption of Antarctic krill by minke whales. Antarctic Science 3(1): 13-8. Food web models and data for NW African ecosystem, L. Morissette et al.  48 Barlow, J. 2003a. Preliminary estimates of the abundance of cetaceans along the U.S. west coast: 1991-2001. Administrative Report, Admin. Rept. LJ-03-03. Southwest Fisheries Science Center (SWFSC), National Marine Fisheries Service, (NMFS), National Oceanic and Atmospheric Administration (NOAA), La Jolla, CA, USA. Barlow, J. 2003b. Cetacean abundance in Hawaiian waters during summer/fall 2002. Administration report, Admin. Rept. LJ-03- 13. Southwest Fisheries Science Center (SWFSC), National Marine Fisheries Service (NMFS), National Oceanic and Atmospheric Administration (NOAA), La Jolla, CA, USA. Barlow, J. 2006. Cetacean abundance in Hawaiian waters estimated from a summer/fall survey in 2002. Marine Mammal Science, 22: 446-464. Barlow, J. and Taylor, B.L. 2005. Estimates of sperm whale abundance in the northeastern temperate Pacific from a combined acoustic and visual survey. Marine Mammal Science, 21: 429-445. Bas, C. 1993. Long-term Variability in the Food Chains, Biomass Yields, and Oceanography of the Canary Current Ecosystem. pp. 94- 103 In Sherman, K., Alesander, D.G., Gold, B.D. (Eds.), Large Marine Ecosystems: Stress, Mitigation, and Sustainability. American Association for the Advancement of Science, Washington DC. Baum, J.K., Myers, R.A., Kehler, D.G., Worm, B., Harley, J. and Doherty, P.A. 2003. Collapse and conservation on shark populations in the Northwest Atlantic. Science 299: 389-392. Bax, N.J. 1998. The significance and prediction of predation in marine fisheries. ICES Journal of Marine Sciences 55: 997-1030. Best, P.B. 1999. Food and feeding of sperm whales Physeter macrocephalus of the west coast of South Africa. South African Journal of Marine Science 21: 393–413. Best, P.B. 2001. Distribution and population separation of Bryde's whale Balaenoptera edeni off southern Africa. Marine Ecology Progress Series 220: 277-289. Best, P.B., Sekiguchi, K., Rakotonirina, B., and Rossouw, A. 1996. The distribution and abundance of humpback whales off southern Madagascar, August-September 1994. Reports of the International Whaling Commission, 46: 323. Best, P.B., Rademeyer, R.A., Burton, C., Ljungblad, D.K., Sekiguchi, H., Shimada, D., Thiele, D., Reeb, D., and Butterworth, D.S. 2003. The abundance of blue whales on the Madagascar Plateau, December 1996. Journal of Cetacean Research and Management 5: 253–260. Bransletter, S. 1987. Age and growth validation of newborn sharks in laboratory aquaria with comments on the life story of the Atlantic sharpnose shark. Copeia 2: 291-299. Brodie, P.F. 1975. Cetacean energetics, an overview of intraspecific size variation. Ecology 56: 152-161 Brown, S.G. and Lockyer, C.H. 1984. Whales. pp. 717-781 In Laws, R.M. (Ed.) Antarctic ecology Vol. 2. Academic Press, London. Bucke, D. 1989. Histology. pp. 69-97 In Austin, B., Austin, D.A. (Eds.), Methods for the Microbiological Examination of Fish and Shellfish. Wiley, New York. Bundy, A., Lilly, G., and Shelton, P. 2000 A mass balance model of the Newfoundland-Labrador shelf. Canadian Technical Report of Fisheries and Aquatic Sciences 2310. Calambokidis, J. and Barlow, J. 2004. Abundance of blue and humpback whales in the eastern North Pacific estimated by capture- recapture and line-transect methods. Marine Mammal Science 20: 63-85. Camarena Luhrs, T. 1986. Les principales espèces de poissons pélagiques côtiers au Sénégal : biologie et évaluation des ressources. Thèse de Doctorat, Université de Bretagne Occidentale, France. 187 p. Caverivière, A. and Thiam, M. 1992. Indices d’abondances et niveaux d’exploitation des espèces démersales du plateau continental Sénégalais. Estimations à partir des résultats des campagnes de chalutage stratifié (1986-1991). CRODT. Document Scientifique n°132. Cayré P., Kothias, J.M., Stretta M., and Diouf, T. 1988. La biologie des thonidés de l’Atlantique tropical Est, pp. 157-264 In A. Fonteneau, A. and Marcille, J. (Eds.) Ressources, pêche et biologie des thonidés tropicaux de l’Atlantique tropicale Centre Est. FAO. Document Technique des Pêches No. 292. Chassot E., Balguerias E., Guitton J., Jouffre D., Tfeil B. and Gascuel D. 2006. Diagnostic de l’état du stock de poulpe (Octopus vulgaris) mauritanien. Synthèse et nouvelles évaluations par approche globale. Rapport de la sous commission Poulpe, groupe d’évaluation des stocks démersaux, IMROP, Nouadhibou 11-16 janvier 2006. Christensen, V. and Pauly, D. 1992. The ECOPATH II – 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., Walters, C.J., and Pauly, D. 2000. Ecopath with Ecosim: a User’s Guide. October 2000 Edition, Fisheries Centre, University of British Columbia, Vancouver, B.C. and ICLARM, Penang, Malaysia. 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. Clark, W.G. 1982. Historical rates of recruitment to Southern Hemisphere fin whale stocks. Report of the International Whaling Commission 32:305–324. Clarke, M.R., Martins, H.R. and Pascoe, P. 1993. The diet of sperm whale (Physeter macrocephalus Linnaeus 1758) off the Azores. Philosophical Transactions of the Royal Society of London (Series B) 339: 67-82. Corkeron, P.J. and Connor, R.C. 1999. Why do baleen whales migrate? Marine Mammal Science 15: 1228–1245 Crawford, R.J.M., P.G. Ryan 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. Currie, P. 2007. Whales, Sustainability and International Environmental Governance. Review of European Community & International Environmental Law 16: 45-57 Davis, R.W. and Fargion, G.S. 1996. Distribution and Abundance of Marine Mammals in North-Central and Western Gulf of Mexico: Final Report. Volume II: Technical Report. Technical Report, OCS Study MMS 96-0027. US Department of the Interior, Minerals Management Service. Davis, R.W., Evans, W.E., and Wursig, B. 2000. Cetaceans, Sea Turtles and Seabirds in the Northern Gulf of Mexico: Distribution, Abundance and Habitat Associations. Volume II: Technical Report. Technical Report, OCS Study MMS 2000-003 - USGS/BRD/CR-1999-0005. Texas A&M University at Galveston and the National Marine Fisheries Service. De la Cruz-Aguero, G. 1993. A preliminary model of Mandinga lagoon, Veracruz, Mexico, pp. 193-196. In Christensen, V. and Pauly, D. (Eds.) Trophic models of aquatic ecosystems. ICLARM Conference Proceedings 26. De la Mare, W.K. 1985. On the estimation of mortality rates from whale age data, with particular reference to minke whales (Balaenoptera acutorostrata) in the Southern Hemisphere. Report of the International Whaling Commission 35: 239-250. Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  49 Diallo, I., Cissé, I., and Bah, A. 2004. Modèle trophique du système côtier du plateau continental Guinéen, pp. 113-123. In Palomares, M.L.D. and Pauly, D. (Eds.) West African marine ecosystems: models and fisheries impacts. Fisheries Centre Research Reports 12(7). Diatta, Y., Ba, F.L., and Capape, C. 2001. Rôle trophique du poulpe commun, Octopus vulgaris, chez les élasmobranches de la côte du Sénégal (Atlantique oriental tropical); comparaison avec les espèces des côtes Tunisiennes (Méditerranée centrale). Acta Adriat. 42 (1): 77-88. Diouf, T. 1980. Pêche et biologie de trois scombridae exploités au Sénégal: Euthynnus alletteratus, Sarda sarda et Scomberomorus tritor. Thèse de Doctorat 3ème cycle, Université de Bretagne Occidentale, France. 159 p. Essington, T.E., 2006. Pelagic Ecosystem Response to a Century of Commercial Fishing and Whaling. pp. 38-49 In Estes, J.A., DeMaster, D.P., Doak, D.F., Williams, T.M., and Brownell Jr R.L. (Eds.) Whales, Whaling, and Ocean Ecosystems. University of California Press, Berkeley CA. Estes, J.A., DeMaster, D.P., Doak, D.F., Williams, T.M. and Brownell Jr., R.L. 2007. Whales, whaling, and ocean ecosystems. University of California Press, 418 p. Evans, P.G.H. 1998. Biology of cetaceans of the north-east Atlantic (in relation to seismic energy). Proceedings of the Seismic and Marine Mammals Workshop, London, 23-25 June 1998 (Eds.) Mark L Tasker and Caroline Weir. FAO, 2002. Report of the FAO Working Group on the Assessment of Small Pelagic Fish off Northwest Africa - Banjul, Republic of the Gambia, 5-12 April 2002. FAO Fisheries Report No. 686. FAO. 2003. Report of the FAO Working Group on the Assessment of Small Pelagic Fish off Northwest Africa - Agadir, Morocco, 31 March-10 April 2003. FAO Fisheries Report No. 723. FAO CECAF Scientific Committee. 2002. Status of stocks and resources - Sardine - Northwest of Africa, 2002. FIRMS Reports. FAO - Rome. Updated Mar 20 2006. Available via FIRMS-Web from: http://firms.fao.org/firms/resource/10086 [Accessed May 23 2008]. Fedoseev, A., 1970. Geostrophic circulation of surface waters on the shelf of north-west Africa. Rapp. P.-V. Reun. Cons. Int. Explor. Mer., 159, 32-37. Fernández, L., Salmerón, F. and Ramos, A. 2005. Change in Elasmobranchs and Other Incidental Species in the Spanish Deepwater Black Hake Trawl Fishery off Mauritania (1992–2001). J. Northw. Atl. Fish. Sci., 35: 325-331. Findlay, K.P., Best, P.B., Peddemors, V.M., and Gove, D. 1994. The distribution and abundance of humpback whales on their southern and central Mozambique winter grounds. Reports of the International Whaling Commission 44: 311-320. Fischer, W., G. Bianchi and Scott, W.B. 1981. Fiches FAO d’identification des espèces pour les besoins de la pêche. Atlantique centre–est ; zones de pêches 34, 47 (en partie). Canada Fonds de dépôts. Ottawa, Ministère des Pêcheries et Océans, en accord avec la FAO, Vols. 1-7. Forney, K.A. and Barlow, J. 1993. Preliminary winter abundance estimates for cetaceans along the California coast based on a 1991 aerial survey. Reports of the International Whaling Commission, 407-415. Forney, K.A., Barlow, J., and Carretta, J.V. 1995. The abundance of cetaceans in California waters: Part II. Aerial surveys in winter and spring of 1991 and 1992. US National Marine Fisheries Service Fishery Bulletin, 93, 15-26. Forcada, J., Notarbartolo, D.S.G., and Fabbri, F. 1995. Abundance of fin whales and striped dolphins summering in the Corso- Ligurian Basin. Mammalia 59: 127-140. Franqueville, C. 1983. Biologie et dynamique de population des daurades (Pagellus bellottii, Steindacher, 1982) le long des côtes Sénégalaises. Thèse Doc. D’Etat. Université d’Aix-Marseille II. 276 p. Fréon, P. 1988. Réponses et adaptation des stocks de clupéidés d’Afrique de l’Ouest à la variabilité du milieu ; analyse et réflexion à partir de l’exemple du Sénégal. Études et thèses, ORSTOM, Paris. Garrison, L.P., Swartz, S., Martinez, A., Burks, C., and Stamates, J. 2003. A marine mammal assessment survey of the Southeast US Continental Shelf: February - April 2002. NOAA Technical Memorandum, NMFS-SEFSC-492. U.S. Department of Commerce. Gascuel D. and Laurans, M. 2003. Évaluation  des stocks par l'approche globale et évolutions d'abondance. Synthèse du groupe de travail "Analyses mono-spécifiques" de Mindelo (octobre 2001). pp. 35-40 In Gascuel D., Barry M., Laurans M., and Sidibé A. 2003. Évaluations des stocks démersaux en Afrique du Nord-Ouest, travaux du groupe "Analyses mono-spécifiques" du projet SIAP. COPACE / PACE, 03/65, Rome FAO. Gascuel, D. Labrosse, P., Meissa, B. Taleb Sidi, M.O. and Guénette, S. 2007. Decline of demersal resources in North-West Africa: an analysis of Mauritania trawl-survey data over the past 25 years. African Journal of Marine Science 29(3):  331-345. Gerrodette, T. and Forcada, J. 2002. Estimates of abundance of western/southern spotted, whitebelly spinner, striped and common dolphins, and pilot, sperm and Bryde's whales in the Estern Tropical Pacific Ocean. Administrative Report, Admin. Rept. LJ- 02-20. Southwest Fisheries Science Center (SWFSC), National Marine Fisheries Service, (NMFS), National Oceanic and Atmospheric Administration (NOAA), La Jolla, California. Gonzalez, A.F., Lopez, A., Guerra, A., Barreiro, A. 1994. Diets of marine mammals on the northwestern Spanish Atlantic coast with special reference to Cephalopoda. Fisheries Research 21: 179-191 Guénette, S. and Christensen, V. 2005. Food web models and data for studying fisheries and environmental impacts in Eastern Pacific ecosystems. Fisheries Centre Research Reports 13(1). Guénette, S., Heymans, S.J.J., Christensen, V., and Trites, A.W. 2006. Ecosystem models show combined effects of fishing, predation, competition, and ocean productivity on Steller sea lions (Eumetopias jubatus) in Alaska. Canadian Journal of Fisheries and Aquatic Sciences 63: 2495-2517. Hammill, M.O. and Stenson, G.B. 2007. Application of the precautionary approach and conservation reference points to management of Atlantic seals. ICES Journal of Marine Science 64: 702–706. Hansen, L.J., Mullin, K.D., and Roden, C.L. 1995. Estimates of cetacean abundance in the northern Gulf of Mexico from vessel surveys, Contribution No. MIA-94/95-25. SEFSC, Miami Laboratory, Miami, USA. Haug, T., Gjøsæter, H., Lindstrøm, U. and Nilssen, K.T. 1995. Diet and food availability for Northeast Atlantic minke whales (Balaenoptera acutorostrata), during the summer of 1992. ICES Journal of Marine Science 52: 77-86. Haug, T., Lindstrøm, U., Nilssen, K.T., Røttingen, I. and Skaug, H.J. 1996. Diet and food availability for Northeast Atlantic minke whales, Balaenoptera acutorostrata. Report of the International Whaling Commission 46: 371-82. Heymans, J.J. 2005. Ecosystem model of the Eastern Aleutians and the Gulf of Alaska in 1963. pp 83-105 In Guénette, S. and Christensen, V. (Eds.). Food web models and data for studying fisheries and environmental impacts on Eastern Pacific ecosystems. Fisheries Centre Research Reports 13(1). Holt, S.J. 2007. Whaling: Will the Phoenix rise again? Marine Pollution Bulletin 54: 1081-1086 Food web models and data for NW African ecosystem, L. Morissette et al.  50 ICCAT. 1999a. Rapport de la période biennale 1998-1999. 1ère partie (1998)-Vol. 1. Version Française. Madrid, Espagne. ICCAT. 1999b. Rapport de la période biennale 1998-1999. 1ère partie (1998)-Vol. 2. Version Française. Madrid, Espagne. ICCAT. 1999c. Report of the ICCAT SCRS skipjack stock assessment session, Funchal, Madeira, Portugal, 28 June to 2 July 1999, ICCAT Report SCRS/99/21, 52 pp. ICCAT. 2006. Report on the Standing Committee on Research and Statistics (SCRS), Madrid, Spain, October 2 to 6, 2006. 195 pp. Innes S., Lavigne, D.M., Eagle, W.M., Kovacs, K.M. 1986. Estimating feeding rates of marine mammals from heart mass to body mass ratios. Marine Mammal Science 2: 227-229. Jackson, J.B.C. 2007. When ecological pyramids were upside down. pp. 23-37 In Estes, J.A., DeMaster, D.P., Doak, D.F., Williams, T.M. and Brownell Jr., R.S. (Eds.) Whales, Whaling, and Ocean Ecosystems. University of California Press, Berkeley and Los Angeles, California. Jackson, J. B. C., M.X. Kirby, W. H. Berger, K.A: Bjorndal, L. W. Botsford, B. J. Bourque, R. H. Bradbury, R. Cooke, J. Erlandson, J. A. Estes, T. P. Hughes, S. Kidwell, C. B. Lange, H. S. Lenihan, J. M. Pondolfi, C. H. Peterson, R. S. Steneck, M. J. Tegner, and R. R. Warner. 2001. Historical overfishing and the recent collapse of coastal ecosystems. Science 293: 629-638. Jarre-Teichmann, A. 1998. The potential role of mass balance models for the management of upwelling ecosystems. Ecological Applications/Supplement 8(1): 93-103. 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. Kaschner, K. 2004. Modelling and mapping of resource overlap between marine mammals and fisheries on a global scale. Ph.D., University of British Columbia. Katona, S. and Whitehead, H. 1988. Are cetaceans ecologically important? Oceanography and Marine Biology Annual Reviews 26: 553-568. Kawakami, T. 1980. A review of sperm whale food. Scientific Report of the Whales Research Institute Tokyo 32: 199-218. Klinowska, M. 1991. Dolphins, porpoises, and whales of the world: the IUCN red data book. IUCN - The World Conservation Union, Gland, Switzerland. Laurans, M. 2005. Évaluation des ressources halieutiques en Afrique de l'ouest : dynamique des populations et variabilité écologique. Thèse de Doctorat Halieutique, Ensa-Rennes. Organisme d'accueil : Pôle Halieutique Agrocampus Rennes. Lockyer, C. 1981. Growth and energy budgets of large baleen whales from the southern hemisphere. p. 379-488. In FAO Fisheries Series 5 ; MammaIs in the Sea. Lydersen, C., Weslawski, J.M., and Oritsland, N.A. 1991. Stomach content analysis of minke whales Balaenoptera acutorostrata from the Lofoten and Vesteralen areas, Norway. Holarctic Ecology, 14, 219-222. Mackinson, S., Blanchard, J.L., Pinnegar, J.K., and Scott, R. 2003. Consequences of alternative functional response formulations in models exploring whale-fishery interactions. Marine Mammal Science 19: 661-681. Maxim, C. 1995. Horse mackerel and false scad stock assessment and catch projections, CECAF Division 34.1.3 and 34.3.1. Sci. Mar. 59(3-4): 611-627. Medina-Gaertner, M. 1985. Étude du zooplancton côtier de la Baie de Dakar et de son utilisation par les poissons comme source de nourriture. Thèse, Université de Bretagne Occidentale. 141 p. Mitchell, E. 1973. Draft report on humpback whales taken under specific scientific permit by eastern Canadian land stations, 1969- 1971. Report of the International Whaling Commission, 23: 138-154. Mobley, J.R.J., Spitz, S.S., Forney, K.A., Grotefendt, R.A., and Forestall, P.H. 2000. Distribution and abundance of odontocete species in Hawaiian waters: Preliminary results of 1993-98 aerial surveys, Administration Report LJ-00-14C. Southwest Fisheries Science Center, National Marine Fisheries Service, La Jolla, California, USA. Mohammed, E. 2003. A generic marine ecosystem model for the Southeastern Caribbean in the late 1990s: application to Grenada and the Grenadines. pp. 191-225 In Zeller, D., Booth, S., Mohammed, E., and Pauly, D. (Eds.) From Mexico to Brazil: Central Atlantic fisheries catch trends and ecosystem models. Fisheries Centre Research Reports 11(6). Morissette, L. 2005. Addressing uncertainty in ecosystem modelling. Pages 127-142 In Levner, E, Linkov, I. and Proth, J.M. (eds.). Strategic management of marine ecosystems. NATO Science Series: IV: Earth and environmental sciences, Volume 50. Springer-Verlag, Heidelberg, Germany. Morissette, L. 2007. Complexity, cost and quality of ecosystem models and their impact on resilience: a comparative analysis, with emphasis on marine mammals and the Gulf of St. Lawrence. PhD thesis, Zoology, University of British Columbia, Vancouver BC, Canada. Morissette, L., Despatie, S.-P., Savenkoff, C Hammill, M.O., Bourdages, H. and Chabot, D. 2003. Data gathering and input parameters to construct ecosystem models for the northern Gulf of St. Lawrence (mid-1980s). Canadian Technical Report of Fisheries and Aquatic Sciences. 2497: vi+94 p. Mullin, K.D. and Fulling, G.L. 2003. Abundance of cetaceans in the southern U.S. Atlantic ocean during summer 1998. Fishery Bulletin 101: 603-613. Mullin, K.D. and Fulling, G.L. 2004. Abundance of cetaceans in the oceanic northern Gulf of Mexico, 1996-2001. Marine Mammal Science 20: 787–807. Myers, R. A. and Worm, B. 2003. Rapid worldwide depletion of predatory fish communities. Nature 423: 280-283. Nieland, H. 1982. The food of Sardinella aurita (Val.) and Sardinella eba (Val.) off the coast of Senegal. Rapp. P.-V. Réun. J. Cons. int. Explor. Mer 180: 369-373. Nordøy, E.S. and Blix, A.S. 1992. Diet of minke whales in the northeastern Atlantic. Report of the International Whaling Commission 43: 393–399. Northridge, S.P. 1984. World review of interactions between marine mammals and fisheries. FAO Fisheries Technical Paper, No.251. 197 p. Okey, T.A. 2001. A ‘straw-man’ Ecopath model of the Middle Atlantic Bight continental shelf, United States. pp. 151-166 In Guénette, S., Christensen, V., and D. Pauly (eds.). Fisheries impacts on North Atlantic ecosystems: Models and analyses. Fisheries Centre Research Reports 9(4). Okey, T.A. 2004. Shifted community states in four marine ecosystems: some potential mechanisms. Ph.D. Thesis, Zoology, University of British Columbia, Vancouver, BC, Canada. Okey, T. A. and Pauly, D. 1999. A trophic mass-balance model of Alaska's Prince William Sound ecosystem, for the post-spill period 1994–1996. Fisheries Centre Research Report 7, 155pp Olsen, E. and Holst, J. C. 2001. A note on common minke whale (Balaenoptera acutorostrata) diets in the Norwegian Sea and the North Sea. Journal of Cetacean Research and Management, 3: 179-183. Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  51 Opitz, S. 1993. A quantitative model of the trophic interactions in a Caribbean coral reef ecosystem, p. 259-268. In Christensen, V. and Pauly, D. (Eds.) Trophic models of aquatic ecosystems. ICLARM Conf. Proc. 26. Paine, R.T., Tegner, M.J., and Johnson, E.A. 1998. Compounded perturbations yield ecological surprises. Ecosystems 1: 535-545. Palomares, M.L.D. and Pauly, D. 1998. Predicting the food consumption of fish populations as functions of mortality, food type, morphometrics, temperature and salinity. Marine and Freshwater Research 49(5): 447-453. Paula e Silva, R. de, R., M.I. Sousa and A. M. Caramelo. 1993. The Maputo Bay ecosystem (Mozambique), pp. 214-223 In Christensen, V and Pauly, D. (Eds.) Trophic models of aquatic ecosystems. ICLARM Conf. Proc. 26. ICLARM, Manila, Philippines. Pauly, D. 1978. A preliminary compilation of fish length growth parameters. Ber. Inst. Meereskd. Christian-Albrechts-Univ. Kiel (55):1-200. Pauly, D. 1979. Gill size and temperature as governing factors in fish growth: a generalization of von Bertalanffy’s growth formula. Ber. Inst. Meereskd. Christian-Albrechts Univ. Kiel 63. 156 p. Pauly, D. 1980. On the interrelationships between natural mortality, growth parameters, and mean environmental temperature in 175 fish stocks. J. Cons. int. Explor. Mer 39(2): 175-192. Pauly, D. and Munro, J. 1984. Once more on growth comparison in fish and invertebrates. Fishbyte 2(1):21. Pauly, D. Christensen, V. and Haggan, N. (Eds.). 1996. Mass-balance model of Alaska Gyre. Fisheries Centre Research Reports 4(1). Pauly, D., Christensen, V., Guénette, S. Pitcher, T.J., Sumaila, U.R., Walters, C.J., Watson, R. and Zeller, D. 2002. Towards sustainability in world fisheries. Nature 418: 689-695. Perry, S.L., DeMaster, D.P., and Silber, G.K. 1999. The great whales: History and status of six species listed as endangered under the U.S. Endangered Species Act of 1973. Marine Fisheries Review, 61(1): 1-74. Plagányi, É.E. 2007. Models for an ecosystem approach to fisheries. FAO fisheries technical paper 477. 126 p. Plagányi, E.E. and D.S. Butterworth. 2002. Competition with fisheries. pp 268-273 In Perrin, W.F., Würsig, B., and Thewissen, H.G.M. (Eds.) Encyclopedia of Marine Mammals. Academic Press, San Diego. Plagányi, É.E. and D.S. Butterworth. 2004. A critical look at the potential of Ecopath with Ecosim to assist in practical fisheries management. African Journal of Marine Sciences 26 : 261-287. Postel, E. 1955. Résumé des connaissances acquises sur les Clupéidés de l’Ouest Africain. Rapp. J. Cons. int. Explor. Mer 137: 14-17. Punt, A.E. and Butterworth, D.S. 1995. The effects of future consumption by the Cape fur seal on catches and catch rates of the Cape hakes. 4. Modelling the biological interaction between Cape fur seals Arctocephalus pusillus pusillus and the Cape hakes Merluccius capensis and M. paradoxus. South African Journal of Marine Science 16: 255-285. Roberts, S.M. 2003. Examination of the stomach contents from a Mediterranean sperm whale found south of Crete, Greece. Journal of the Marine Biological Association of the United Kingdom 83:667-670. Rosenbaum, H.C., Strindberg, S., and Ersts, P.J. 2004. Initial estimates of abundance and distribution of humpback whales on their wintering grounds in the coastal waters of Gabon (southeastern Atlantic Ocean, Area B) based on aerial surveys. (SC/56/SH2). In International Whaling Commission - Scientific Committee Meeting, pp. 9. (unpublished), Sorento, Italy. Ruesink, J. L. 1998. Ecosystem models as caricatures: the case of Prince William Sound. Pp. 87-89 In Okay, T.A. and Pauly, D. (Eds.) Trophic mass-balance model of Alaska's Prince William Sound ecosystem for the post-spill period 1994–1996. Fisheries Centre Research Report 7. Saetersdal, G., Ofstad, A., Nore, A., Olsen, S., and Moore G.K.F. 1995. Survey of the pelagic fish resources off North West Africa. Senegal-Gambia. 7-16 November 1995. NORAD/FAO/UNDP GLO 82/001. 16 p Samb, B. 1988. Seasonal growth, mortality, recruitment pattern of Sardinella maderensis off Senegal, pp. 257-271 In Venema, S.C., Christensen, J.M., and Pauly, D. (Eds.) Contributions to tropical fisheries biology. FAO/DANIDA Follow-up Training Course on Fish Stock Assessment in the Tropics, Denmark 1986, Philippines, 1987. FAO Fisheries Report n° 389. Rome. Samb, B. 1997. Synthèse des évaluations récentes sur l’état des stocks de sardinelles, chinchards et maquereaux dans la zone nord COPACE. Document COPACE/FAO. 11th Session du COPACE/FAO sur l’Evaluation des Stocks.24 au 26 Septembre 1997, Accra, Ghana. Samb, B. and Mendy, A.N. 2004. Dynamique du réseau trophique de l’écosystème Sénégambien en 1990, pp. 57-70 In Palomares, M.L.D. and Pauly, D. (Eds) West African marine ecosystems: models and fisheries impacts. Fisheries Centre Research Reports 12(7). Fisheries Centre, UBC, Vancouver. Samb, B. and Pauly, D. 2000. On ‘variability’ as a sampling artifact: the case of Sardinella in north-western Africa. Fish and Fisheries 1: 206-210. Santos, M.B., Martin, V., Arbelo, M., Fernandez, A. and Pierce, G.J. 2007. Insights into the diet of beaked whales from the atypical mass stranding in the Canary Islands in September 2002. Journal of Mar. Ass. U.K. 87: 243-251. Scheffer, M., S. Carpenter, J. A., Foley, C. Folke, and Walker, B. 2001. Catastrophic shifts in ecosystems. Nature 413: 591-596. Schneider, W. 1990. FAO species identification sheets for fishery purposes. Field guide to the commercial marine resources of the Gulf of Guinea. Prepared and published with the support of the FAO Regional Office for Africa. FAO, Rome. 26 p. Sergeant, D.E. 1977. Stocks of fin whales Balaenoptera physalus in the North Atlantic Ocean. Report of the International Whaling Commission 27: 460–473. Sidi, M.T. and Guénette, S. 2004. Modèle trophique de la ZEE mauritanienne: comparaison de deux périodes (1987 et 1998). pp 12- 38 In Palomares, M.L.D. and Pauly, D. (Eds.) West African marine ecosystems : models and fisheries impacts. Fisheries Centre Research Reports 12(7). Sidibé, A. 2003. Les ressources halieutiques démersales côtières de la guinée ; exploitation, biologie et dynamique des principales espèces de la communauté à Sciaenidés. Ph.D. Thesis, Ecole Nationale Supérieure Agronomique de Rennes, France. Sigurjónsson, J. and Víkingsson, G.A. 1997. Seasonal abundance of and estimated food consumption by cetaceans in Icelandic and adjacent waters. Journal of Northwest Atlantic Fishery Science, 22: 271-287. Sivertsen K. 2006. Overgrazing of kelp beds along the coast of Norway. Journal of Applied Phycology. 18. 599-610. Sivertsen, S.P., Pedersen, T., Lindstrøm, U.L.F and Haug, T. 2006. Prey partitioning between cod (Gadus morhua) and minke whale (Balaenoptera acutorostrata) in the Barents Sea. Marine Biology Research, 2(2): 89-99. Smout, S. and Lindstrom, U. 2007. Multispecies functional response of the minke whale Balaenoptera acutorostrata based on small-scale foraging studies. Marine Ecology Progress Series 341: 277-291. Springer, A.M., Estes, J.A., van Vliet, G.B., Williams, T.M., Doak, D.F., Danner, E.M., Forney, K.A., and Pfister, B. 2003. Sequential megafaunal collapse in the North Pacific Ocean; an ongoing legacy of industrial whaling? Proceedings of the National Academy of Sciences 100: 12 223-12 228. Stanford, R., Lunn, K., and Guénette, S. 2001. A preliminary ecosystem model for the Atlantic coast of Morocco in the mid-1980s. pp Food web models and data for NW African ecosystem, L. Morissette et al.  52 314-344 In Guénette, S., Christensen, V., and Pauly, D (Eds.). Fisheries impacts on North Atlantic Ecosystems: models and analyses. Fisheries Centre Research Reports 9(4). Stewart, D.A., Agnew, D., Boyd, R., Briggs, R., and Toland, P. 1993. The derivation of changes in Nephrops per unit effort values for the Northern Ireland fishing fleet. Fisheries Research 17, 273-292. Stobberup, K.A., Ramos, V.D.M., and Coelho, M.L. 2004. Ecopath model of the Cape Verde coastal ecosystem, pp. 39-56 In Palomares, M.L.D. and Pauly, D. (Eds.) West African marine ecosystems: models and fisheries impacts. Fisheries Centre Research Reports 12(7). Fisheries Centre, UBC, Vancouver. Struck, D. 2001. Japan blames whales for lower fish catch. International Herald Tribune (July 28-29). Tamura, T. 2003. Regional assessments of prey consumption and competition by marine cetaceans in the world. In Responsible Fisheries in the Marine Ecosystem. pp. 143-170 In Sinclair, M. and Valdimarsson, G. (Eds). Responsible Fisheries in the Marine Ecosystem, Fishery Industries Division, CAB International: FAO, Rome (Italy) and Wallingford (United Kingdom). Toresen, R. 1996. Survey of the pelagic fish resources off North West Africa. Part 3 Senegal-Gambia. 1-9 November 1996. NORAD/FAO/UNDP GLO 92/013. 13 p. Toresen, R. 1997. Survey of the pelagic fish resources off North West Africa. Cruise report N° 11/97. 4 - 12 November 1997. Part I Senegal-Gambia. NORAD/FAO/UNDP GLO 92/013. 14 p. Toresen, R. 1998. Survey of the pelagic fish resources off North West Africa. Cruise report N° 11/98. 30 October - 9 November 1998. Part I Senegal-Gambia. NORAD/FAO/UNDP. 14 p. Toresen, R. and Kolding, J. 1999. Survey of the pelagic fish resources off North West Africa. Cruise report N° 11/98. 30 October - 9 November 1999. Part I Senegal-Gambia. NORAD/FAO/UNDP. 31 p. Touré, D. 1983. Contribution à l’étude de l’upwelling de la baie de Gorée (Dakar Sénégal) et de ses conséquences sur le développement de la biomasse phytoplanctonique. Doc. Sci. CRODT n° 93. Trites, A.W. 2002. Predator-prey relationships. pp 994-997 In Perrin, W.F., Wursig, B., and H.G.M. Thewissen (eds.). Encyclopedia of Marine Mammals. Academic Press, San Diego. Trites, A.W. and Pauly, D. 1998. Estimating mean body masses of marine mammals from maximum body lengths. Canadian Journal of Zoology 76: 886-896. Trites, A.W. and Heise, K. 1996. Marine Mammals: Southern BC Shelf Model, pp. 51-55. In: D. Pauly, V. Christensen and N. Haggan (eds.) Mass balance models of the north-eastern Pacific ecosystems. Fisheries Centre Research Reports 4(1), University of British Columbia, Vancouver. Vasconcellos, V., Mackinson, S., Sloman, K., and Pauly, D. 1997. The stability of trophic mass-balance models of marine ecosystems: a comparative analysis. Ecological Modelling 100: 125-143. Voituriez, B. and Herbland, A.. 1982. Comparaison des systèmes productifs de l’Atlantique Tropical Est : dômes thermiques, upwellings côtiers et upwellings équatorial. Rapp. P.-V. Réun. CIEM, 180: 114-30. Wade, P.R. and Gerrodette, T. 1993. Estimates of cetacean abundance and distribution in the Eastern Tropical Pacific. Reports of the International Whaling Commission 43: 477-493. Walters, C. and Martell, S. 2004. Fisheries ecology and management. Princeton University Press, Princeton, New Jersey. Witteveen, B.H., Foy, R.J. and Wynne, K.M. 2006. The effect of predation (current and historical) by humpback whales (Megaptera novaeangliae) on fish abundance near Kodiak Island, Alaska. Fish. Bull. 104 : 10-20. Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  53 FOOD WEB MODEL AND DATA FOR STUDYING THE INTERACTIONS BETWEEN MARINE MAMMALS AND FISHERIES IN THE CARIBBEAN ECOSYSTEMa Jenny Lynn Melgo1, Lyne Morissette1,2,3, Kristin Kaschner4 and Leah R. Gerber1  1Arizona State University, School of Life Sciences, P.O. Box 874501, Tempe, AZ 85287-4501, USA 2Institut des sciences de la mer de Rimouski, 310, Allée des Ursulines, C.P. 3300, Rimouski, QC, G5L 3A1, CANADA 3Fisheries Centre, The University of British Columbia 2202 Main Mall, Vancouver BC V6T 1Z4, CANADA 4Albert-Ludwigs-University, Institute of Biology I (Zoology) Evolutionary Biology and Ecology Lab, Freiburg, GERMANY Lyne.Morissette@globetrotter.net Jennylynn.Melgo@asu.edu Kristin.Kaschner@biologie.uni-freiburg.de Leah.Gerber@asu.edu ABSTRACT A mass balanced ecosystem model was constructed to examine the trophic interaction between whales and fisheries in the Caribbean ecosystem. This model includes data from the ecosystem models of the Lesser Antilles, Bahia Ascencion Mexican Caribbean, Southern Mexican Caribbean, Grenada and the Grenadines, Colombian Caribbean Sea, Caribbean coral reef ecosystem, Costa Rica, Eastern Tropical Pacific and Central Atlantic. Our model includes twenty nine trophic groups for the Caribbean ecosystem model, comprising 10 marine mammal groups, one seabird group, one sea turtle group, eleven fish groups, one cephalopod group, one crustaceans and benthos group, one benthic producers group, one zooplankton, one phytoplankton and one detritus group. Time series catches of some cetaceans and all fish and non-fish groups in 1987’s to 2005’s are also included in the model. Data quality, uncertainty, Ecoranger and simulation analysis for the removal of great whales and fisheries in the ecosystem are included in this report.  INTRODUCTION A food web is defined as the complex trophic links between organisms (prey-predator) living in an ecosystem. The prey-predator relationships in food webs are typically arranged in oriented graphs with hierarchical or layered structures that show energy pathways and matter transfer within the ecosystem (Paine et al. 1998; Pimm 2002). The structure and functions of food webs are  a Cite as: Melgo, J.L., Morissette, L., Kaschner, K., and Gerber, L. (2009) Food web model and data for studying the interactions between marine mammals and fisheries in the Caribbean ecosystem. In: Morissette, L., Melgo, J.L., Kaschner, K. and Gerber, L.R. (eds.) Modelling the trophic role of marine mammals in tropical areas: data requirements, uncertainty, and validation. Fisheries Centre Research Reports XX(X). Fisheries Centre, University of British Columbia, Vancouver, Canada, pp. 48-107.  Food web models and data for the Caribbean model, J.L. Melgo et al.  54 important to many ecologists and biologists in understanding the role of certain species and the ecosystem, predator-prey interactions on how directly or indirectly they influence each other in the ecosystem and the possible competition among species in the area (Trites et al. 1997; Bax 1998; Pauly et al. 1998a; Constable and Gales 2002; Morissette et al. 2006). For example, in the context of marine mammals and fisheries, Estes et al. (2007) assert that the ocean ecosystems throughout the world have experienced a dramatic shift in structure as a result of the removal of top predators and extensive fishing activities. Moreover, when complex trophic interactions are taken into consideration, it has been shown that culling of marine mammals would not lead necessarily to the recovery of fish stocks (Punt and Butterworth 1995; Plagányi and Butterworth 2002; Morissette 2007). The application of food web models to these questions and the generated results may allow ecologists, managers and even policy makers to identify which species or communities might be vulnerable to disturbance and would need immediate attention. Currently, little is known about the interaction of marine mammals and fisheries in the Caribbean ecosystem. The Lesser Antilles pelagic ecosystem model by Mohammed et al. (2007b) investigated the interaction of marine mammals and fisheries in the area. However, their study focuses on marine mammal species that reside almost throughout the year in the waters surrounding the Lesser Antilles, while the present study also includes all of the numerous migratory, large and small marine mammal species occurring in the entire Caribbean ecosystem. There are also other ecosystem models constructed in different areas of the Caribbean Sea and the Gulf of Mexico (Opitz 1996; Wolff et al. 1998; Duarte and Garcia 2002; Mohammed 2003a; Alvarez-Hernández 2003; Vidal and Basurto 2003). These studies focus, however, on the continental shelf, coral reef or coastal and large pelagic components of the ecosystem addressing questions related to the trophic interaction of reef fishes and/or interactions among large pelagic fisheries and some top predators. At least 31 species of marine mammals (6 baleen whales, 24 toothed whales, 1 sirenian) are known to occur in the Caribbean Sea (UNEP 2002). While many of the smaller odontocete species (e.g. dolphins) inhabit this region year-round, most of the larger mysticetes (e.g. humpback whales) migrate to the Caribbean Sea for calving and breeding (Klinowska 1991; Debrot et al. 1998; Reeves et al. 2001; UNEP 2002; Swartz et al. 2003). It has been proposed that some or all of these higher vertebrate species compete with other marine organisms or fisheries for the available food resources (Mohammed et al. 2007b). For example, Tamura (2003) have estimated that marine mammals around the world annually consume 250-440 million tonnes of fish, and thus consume three to six times the amount taken annually by marine fisheries. However, other scientists have shown that there is currently no real scientific evidence for existing large-scale competition between marine mammals and fisheries (Kaschner 2004; Morissette 2007). In addition, there is mounting evidence and a robust documentation of overfishing on a global scale including the Caribbean (Jackson et al. 2001; Pauly et al. 2002; Baum et al. 2003; Myers and Worm 2003; Mahon and McConney 2004), which by itself could easily explain the observed depletion of the fish stocks in this area. In this study, we examined the scientific evidence for the trophic interaction between marine mammals and fisheries (distinguishing between foreign and local fleets); using the local data obtained from several published literatures and reports studied in the adjacent areas of the Caribbean ecosystem supplemented with additional details about marine mammal abundances, diets and food consumption. Our approach was to develop a mass-balanced ecosystem model using the Ecopath and Ecosim (EwE) program (Christensen and Walters 2004). We used the EwE program because it is one of several ecosystem modelling approaches that is widely used in understanding the interactions between marine mammals and fisheries (Morissette 2007). This model allowed us to characterize the structure and functionality of the ecosystem in terms of biomass, mortalities, consumption rates, food habits, general ecosystem indicators and fisheries. Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  55 MATERIALS AND METHODS Study area Our study area was located in the southeastern edge of the Caribbean Sea. It was part of a Large Marine Caribbean Ecosystem including the Dominican Republic at its northern point and extending to Trinidad and Tobago at its southern point. The Caribbean islands covered by our model included Anguilla, Antigua and Barbuda, Barbados, British Virgin Islands, Dominica, Dominican Republic, Grenada and the Grenadines, Guadeloupe, Martinique, Montserrat, Puerto Rico, St Kitts and Nevis, St Lucia, St Vincent, and Trinidad and Tobago, and the US Virgin Island. Generally, the topography of these islands was characterized by steep drop-off with wide pelagic environment, but in some areas (e.g. Virgin Islands, Puerto Rico), we can also observed deep basins and shallow wide areas of coral reefs and seagrass beds (Opitz 1996; UNEP 2002; Martin et al. 2005).  The nominal EEZ’s (within a distance of 200 n. mi. out from its coast) of the Caribbean islands defined the boundary of our study areas, from 10 N to 19 N latitudes and 71.75 W to 56 W longitudes, with a total area of approximately 2 million km2 (Figure 1).   Figure 1. Map of study area: The Caribbean region. The numbers shown in the map were the study sites of eight adjacent South or Central American tropical models which provided data for the confidence interval inputs of each input parameter in our model. Legends:  model study area (green shaded polygon); Bahia Ascencion [1] (Vidal and Basurto 2003); Southern Mexican Caribbean[2] (Alvarez-Hernández 2003); Grenada and the Grenadines[3] (Mohammed 2003); Colombian Caribbean Sea[4] (Duarte and Garcia 2002); Costa Rica[5] (Wolff et al. 1998); Caribbean coral reef[6] (Opitz 1996); Central Atlantic[7] (Vasconcellos and Watson 2004) and Eastern Tropical Pacific[8] (Olson and Watters 2003). Food web models and data for the Caribbean model, J.L. Melgo et al.  56  The pelagic marine environment of the Caribbean Sea is influenced by the North Equatorial currents that pass through between the Lesser Antilles arc (Johns et al. 2002). This water mass inflow is considerably nutrient-poor water which is important for coralline formation in the region (Richards and Bohnsack 1990). The Caribbean Sea is considered to be an ecosystem with low productivity (<150 gCm-2yr-1) based on SeaWiFS global primary productivitiy estimates (Richards and Bohnsack 1990). However, there is considerable spatial and temporal heterogeneity in productivity in some areas (e.g. seagrass beds, mangroves areas, local upwelling, riverine flows) in the region (Heileman 2007). The complex dynamics of these high productive coastal areas and offshore waters contributes and support the Caribbean regions valuable marine ecological and biological diversity (Heileman 2007). Because of its location and unique marine environment, the fishing and tourism industry is an economically important source of income for locals in the region (Opitz 1996; Bacci 1998; CANARI 1999).  MODEL DESCRIPTION The list of trophic groups described in this report is shown in Table 1. This includes whale species, fish species, cephalopods species, crustaceans, planktons, benthic producers and detritus.  Table 1. List of trophic groups and species included in the Ecopath model for Caribbean. Species in bold represent the key species for each of the trophic group based on its importance in biomass and in fisheries in the area. Ecopath group Species 1. Minke whales Balaenoptera acutorostrata 2. Fin whales Balaenoptera physalus 3. Humpback whales Megaptera novaeangliae 4. Bryde’s whales Balaenoptera brydei 5. Sei whales Balaenoptera borealis 6. Blue whales Balaenoptera musculus 7. Sperm whales Physeter macrocephalus, Kogia breviceps, Kogia simus 8. Killer whales Feresa attenuata, Orcinus orca, Pseudorca crassidens 9. Beaked whales Mesoplodon densirostris, Mesoplodon europaeus, Ziphius caviostris 10. Small cetaceans Delphinus capensis, Delphinus delphis, Globicephala macrorhynchus, Grampus griseus, Lagenodelphis hosei, Sousa teuszii, Stenella attenuata, Stenella clymene, Stenella coeruleoalba, Stenella frontalis, Stenella longisrostris, Steno bredanensis, Tursiops truncatus 11. Seabirds Actitis hypoleucos, Ajaia ajaja, Calidris ferruginea, Calonectris diomedea, Ceryle rudis, Chlidonias niger, Halcyon malimbica, Limosa lapponica, Numenius phaeopus, Oceanites oceanicus, Pagrodama nivea, Pelecanus rufescens, Phalacrocorax africanus, Phoenicopterus rubber, Pluvialis squatarola, Sterna caspia, Sterna hirundo 12. Seaturtles Caretta caretta, Chelonia mydas, Eretmochelys imbricata, Dermochelys coriacea 13. Large tunas and billfishes Istiophoridae, Istiophorus albicans, Makaira nigricans, Parexocoetus brachypterus, Tetrapturus albidus, Tetrapturus pfluegeri, Thunnus alalunga, Thunnus albacares, Thunnus obesus, Thunnus thynnus, Xiphias gladius 14. Small tunas Auxis sp., Euthynnus alletteratus, Katsuwonus pelamis, Sarda sarda, Thunnus atlanticus 15. Dolphinfish Coryphaena hippurus, Coryphaenidae 16. Flyingfish Hirundichthys affinis, Cheilopogon cyanopterus, Cypselurus cyanopterus, Parexocoetus brachypterus, Exocoetidae    Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  57 Table 1. (cont.) 17. Other offshore predators Alepocephalidae, Argyropelecus olfersi, Bathylagus nigribenys, Cyclothone sp., Gonostomatidae, Gonosthoma bathyphilum, Lampris guttatus, Lampanyctus macdonaldi, Lobotes surinamensis, Melanocetus sp., Moridae, Myctophidae, Pomatomus saltator, Ruvettus pretiosus, Scopelogadus beanie, Sternoptyx diaphana 18. Pelagic sharks Carcharhinidae, Carcharhinus acronotus, Carcharhinus brevipinna, Carcharhinus falciformis, Carcharhinus leucas, Carcharhinus limbatus, Carcharhinus longimanus, Carcharhinus perezi, Elasmobranchii, Galeocerdo cuvier, Lamnidae, Lamna nasus, Prionace glauca, Isurus oxyrinchus, Isurus paucus, Sphyrna lewini, Urolophidae 19. Coastal sharks and rays Aetobatus narinari, Dasyatidae, Dasyatis americana, Ginglymostoma cirratum, Mustelus sp. , Myliobatidae, Negaprion brevirostris, Raja sp., Rajiformes, rays, Rhizoprionodon terraenovae, small/juvenile sharks, Squalidae 20. Scombrids Acanthocybium solandri, Scomberomorus brasiliensis, Scomberomorus cavalla, Scomberomorus maculatus, Scomberomorus regalis, Scomberomorus sp., Scombridae 21. Small and schooling pelagics Ablennes hians, Alectis ciliaris, Anchoa hepsetus, Anchoa lucida, Anchoa lyolepis, Belonidae, Carangoides ruber, Caranx crysos, Caranx latus, Caranx lugubris, Cetengraulis edentulous, Cetengraulis edentulous, Clupeidae, Clupeiformes, Chloroscombrus chrysurus, Decapterus macarellus, Decapterus punctatus, Decapterus sp., Dorosoma petenense, Elagatis bipinnulata, Elops saurus, Engraulidae, Etrumeus teres, Harengula clupeola, Harengula humeralis, Harengula sp., Hyperoglyphe bythites, Hypoatherina harringtonensis, Jenkinsia lamprotaenia, Lepidocybium flavobrunneum, Neoopisthopterus tropicus, Oligoplites saurus, Opisthonema oglinum, Peprilus alepidotus, Peprilus sp., Platybelone argalus, Sardinella aurita, Sardinella brasiliensis, Selar crumenophthalmus, Selene brevoortii, Selene orstedii, Selene peruviana, Selene setapinnis, Sphyraena barracuda, Sphyraena sp., Stromateidae, Strongylura timucu, Trichiuridae, Trichiurus lepturus, Tylosurus acus, Tylosurus crocodilus  Food web models and data for the Caribbean model, J.L. Melgo et al.  58 Table 1. (cont.) 22. Reef fishes Abudefduf saxatilis, Abudefduf taurus, Acanthostracion polygonius, Acanthostracion quadricornis, Acanthuridae, Acanthurus bahianus, Acanthurus chirurgus, Acanthurus coeruleus, Albula vulpes, Alphestes afer, Aluterus schoepfii, Aluterus scripta, Anisotremus surinamensis, Anisotremus virginicus, Antennarius striatus, Apogon maculates, Aulostomus maculatus, Balistes capriscus, Balistes vetula, Balistidae, Bodianus rufus, Bothus lunatus, Bothus ocellatus, Brotula barbata, Calamus bajonado, Calamus calamus, Calamus pennatula, Cantherhines macrocerus, Cantherhines pullus, Canthidermis sufflamen, Canthigaster rostrata, Centropomus undecimalis, Centropyge argi, Cephalopholis cruentata, Cephalopholis fulva, Chaetodipterus faber, Chaetodon aculeatus, Chaetodon capistratus, Chaetodon ocellatus, Chaetodon sedentarius, Chaetodon striatus, Chilomycterus antennatus, Chilomycterus antillarum, Chromis cyanea, Chromis multilineata, Clepticus parrae, Coryphopterus glaucofraenum, Dactylopterus volitans, Diodon holocanthus, Diodon hystrix, Diplectrum formosum, Diplodus argenteus caudimacula, Echidna catenata, Enchelycore nigricans, Entomacrodus nigricans, Ephippidae, Epinephelus adscensionis, Epinephelus itajara, Epinephelus morio, Epinephelus nigritus, Epinephelus niveatus, Epinephelus striatus, Equetus lanceolatus, Equetus punctatus, Eucinostomus argenteus, Eugerres plumieri, Fistularia tabacaria, Gerreidae, Gerres cinereus, Gnatholepis thompsoni, Gobiosoma evelynae, Gobiosoma horsti, Gramma loreto, Gramma melacara, Gymnothorax funebris, Gymnothorax miliaris, Gymnothorax vicinus, Haemulidae, Haemulon album, Haemulon aurolineatum, Haemulon carbonarium, Haemulon chrysargyreum, Haemulon flavolineatum, Haemulon macrostomum, Haemulon melanurum, Haemulon parra, Haemulon plumieri, Haemulon sciurus, Halichoeres bivittatus, Halichoeres garnoti, Halichoeres maculipinna, Halichoeres poeyi, Halichoeres radiatus, Hemiramphus brasiliensis, Holacanthus ciliaris, Holacanthus tricolor, Hypoplectrus aberrans, Hypoplectrus chlorurus, Hypoplectrus nigricans, Hyporhamphus unifasciatus, Joturus pichardi, Kyphosus incisor, Kyphosus sectatrix, Labridae, Labrisomus guppyi, Labrisomus nuchipinnis, Lachnolaimus maximus, Lactophrys bicaudalis, Lactophrys triqueter, Lopholatilus chamaeleonticeps, Malacanthidae, Malacanthus plumieri, Melichthys niger, Menticirrhus littoralis, Micropogonias furnieri, Micropogonias undulates, Microspathodon chrysurus, Monacanthidae, Monacanthus ciliatus, Monacanthus tuckeri, Mullidae, Mycteroperca bonaci, Mycteroperca rubra, Mycteroperca tigris, Mycteroperca venenosa, Myrichthys breviceps, Myrichthys ocellatus, Myripristis jacobus, Neoniphon marianus, Odontoscion dentex, Ogcocephalus nasutus, Ophidiidae, Ophioblennius atlanticus, Opistognathus aurifrons, Opistognathus macrognathus, Opistognathus maxillosus, Opistognathus whitehursti, Ostraciidae, Parablennius marmoreus, Paralichthyidae, Paralichthys sp., Paranthias furcifer, Pareques acuminatus, Phaeoptyx conklini, Plectrypops retrospinis, Pogonias cromis, Polynemidae, Pomacanthidae, Pomacanthus arcuatus, Pomacanthus paru, Pomacentridae, Priacanthus arenatus, Pseudupeneus maculates, Rypticus saponaceus, Scaridae, Scartella cristata, Scarus coelestinus, Scarus coeruleus, Scarus guacamaia, Scarus iserti, Scarus taeniopterus, Scarus vetula, Scorpaenodes caribbaeus, Serranidae, Serranus tabacarius, Serranus tortugarum, Sparisoma aurofrenatum, Sparisoma chrysopterum, Sparisoma radians, Sparisoma rubripinne, Sparisoma viride, Sphoeroides spengler, Stegastes fuscus, Stegastes leucostictus, Stegastes planifrons, Stegastes variabilis, Stephanolepis setifer, Symphurus chabanaudi, Tetraodontidae, Thalassoma bifasciatum, Xanthichthys ringens, Xyrichtys novacula, Xyrichtys splendens  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  59 Table 1. (cont.) 23. Coastal predators Achirus klunzingeri, Alectis ciliaris, Archosargus probatocephalus, Belonidae, Caranx hippos, Caranx sp., Cynoponticus coniceps, Cynoscion sp., Heteropriacanthus cruentatus, Holocentridae, Holocentrus coruscus, Elegatis bipinnulata, Lutjanidae, Lutjanus analis, Lutjanus apodus, Lutjanus cyanopterus, Lutjanus griseus, Lutjanus jocu, Lutjanus mahogoni, Lutjanus purpureus, Lutjanus synagris, Megalops atlanticus, Mugil cephalus, Mugil curema, Mugil liza, Mugilidae, Mulloidichthys martinicus, Myrichthys breviceps, Myrichthys ocellatus, Myripristis jacobus, Ocyurus chrysurus, Oligoplites sp., Ophichthus ophis, Pempheris poeyi, Peprilus medius, Peprilus snyderi, Pomadasys corvinaeformis, Pomadasys crocro, Rachycentron canadum, Rhomboplites aurorubens, Sargocentron vexillarium, Sciaenidae, Scorpaena grandicornis, Scorpaena inermis, Seriola dumerili, Seriola rivoliana, Seriola sp., Syacium latifrons, Syacium ovale, Synodus foetens, Synodus intermedius, Trachinotus carolinus, Trachinotus falcatus, Trachinotus goodie, Trachinotus paitensis 24. Cephalopods Alloteuthis subulata, Loliginidae, Loligo pealeii, Loligo sp., Loligo vulgaris, Lolliguncula panamensis, Octopodidae, Octopus vulgaris, Ommastrephidae, Sepia bertheloti, Sepia elobyana, Sepia officinalis, Sepia orbignyana, Sepiidae, Thysanoteuthis rhombus, Todarodes sagittatus 25. Crustaceans and benthos Aristeidae, Arca sp., Arcidae, Aristeus antennatus, Aristeus varidens, Bivalvia, Brachyura, Busycon sp., Calappa rubroguttata, Callinectes sapidus, Cancer pagurus, Carcinus maenas, Cardiidae, Cardium edule, Chama crenulata, Chitonidae, Conidae, Crangon crangon, Crangonidae, Crassostrea rhizophorae, Crassostrea virginica, Crepidula porcellana, Crustacea, Diadema sp., Donacidae, Echinoderms, Gastropoda, Geryon maritae, Geryon quinquedens, Haliotis tuberculata, Homarus gammarus, Leucosiidae, Maja squinado, Menippe mercenaria, Miscellaneous marine mollusks, Muricidae, Mytilidae, Naticidae, Necora puber, Nephrops norvegicus, Paguridae, Palaemonidae, Palinurus elephas, Palinurus mauritanicus, Panulirus argus, Panulirus regius, Panulirus sp., Parapanaeus longirostris, Parapenaeopsis atlantica, Pecten maximus, Pectinidae, Penaeidae, Penaeus  kerathurus, Penaeus brasiliensis, Penaeus notialias, Penaeus sp., Perna perna, Pleoticus robustus, Plesionika heterocarpus, Plesiopenaeus edwardsianus, Porifera, Portunidae, Portunus sp., Pyura dura, Ruditapes decussates, Scyllaridae, Solenidae, Strombus sp., Veneridae, Venus rosalina, Venus verrucosa, Volutidae, Xiphopenaeus kroyeri 26. Benthic producers Algae, benthic autotrophs 27. Zooplankton Chaetognatha, Copepoda, Euphausiacea, Hydrozoa, Hyperiidae, Mysidacea, Scyphozoa, ichtyoplankton, macroplankton, meroplankton, planktonic decapods, larvae, and eggs 28. Phytoplankton 29. Detritus  The initial mass-balanced ecosystem model for the Caribbean used herein was originally based on the trophic model of the Lesser Antilles pelagic ecosystem (LAPE) by Mohammed et al. (2007b). There were 31 trophic groups in the initial model, consisting of four marine mammal functional groups, one seabird group, two seaturtle groups, eighteen fish groups, two squid groups, two zooplankton groups, one phytoplankton group and one detritus group. In order to address our goal to examine the trophic interaction between fisheries and whales in the entire Caribbean ecosystem, we modified the structure of this model. In particular, we aggregated some similar non-marine mammal trophic groups based on habitat, feeding category and biological variables (Essington 2006), but retained some important fish species (e.g. dolphinfish, flyingfish) based on local expert suggestions received during the regional workshop held in Barbados on September 2008. We also added an additional 7 cetacean trophic groups to the existing marine mammal groups (e.g. Bryde’s whales, killer whales and deep-diving cetaceans) in the initial model. The modified structure of this model provided a better representation of great whales and of commercially important fish groups in the context of the ‘whales eat fish’ issue. These changes reduced the number of trophic groups of the present model from 31 to 29, consisting of 10 marine Food web models and data for the Caribbean model, J.L. Melgo et al.  60 mammal groups, one seabirds group, one seaturtles group, eleven fish groups, one cephalopods group, one crustaceans and benthos group, one benthic producers group, one zooplankton group, one phytoplankton group, and one detritus group (Table 1). Published literature documenting other Caribbean ecosystem models from the surrounding areas was used to establish confidence intervals inputs for calibration of biomass, P/B, Q/B and diet matrix. Such models were available for Caribbean coral reef (Opitz 1996), Colombian Caribbean Sea (Duarte and Garcia 2002), Southeastern Caribbean (Mohammed 2003a), Southeastern Mexican Caribbean (Alvarez-Hernández, 2003), Bahia Ascencion Mexican Caribbean (Vidal and Basurto 2003), and the coast of Costa Rica (Wolff et al. 1998). In addition, we also considered two models that were not from the Caribbean area, but represented typical pelagic, tuna-dominated ecosystems: the Central Atlantic (Vasconcellos and Watson 2004) and the Eastern tropical Pacific (Olson and Watters 2003) models. These areas were included in the present analysis because they have similar offshore migratory pelagics species as of the Caribbean ecosystem (Mohammed et al. 2007b). In order to maintain the same trophic groups from those eight Ecopath ecosystem models for adjacent Caribbean areas, we matched their existing trophic groups and corresponding species to the structure of our model.  There were some species belonging to the trophic groups in the other Ecopath models that were different from our trophic groups’ species lists. This was resolved by taking the key species of the trophic groups for each ecosystem models and aggregating them according to the structure of our established trophic groups of the present Caribbean model. Diet information of the trophic groups represented the adult diet only, since most of the biomass estimates available were only for the spawning or adult biomass. In general, we incorporated the trophic diet data from Heileman et al. (2007) and Opitz (1996). Cetacean biomass and abundance inputs for all marine mammal groups included in our model were provided by Kristin Kaschner based on a global database of marine mammal occurrence and densities Kaschner 2004) since there were currently no reliable abundance estimates for the majority of  cetaceans from dedicated surveys conducted in the area. Cetacean abundances within the study area were estimated based on global abundance estimates that were converted into densities per 0.5 degree latitude by 0.5 degree longitude cells falling within the range of predicted occurrence for each species and weighted by the relative suitability of the habitat of each cell for a given species (Kaschner et al. 2006). The data provided by Kaschner was updated and validated wherever possible using abundance estimates obtained during dedicated cetacean surveys conducted in other areas associated with similar types of subtropical habitat. Marine mammal mortality rate data were extracted from several published references (e.g. Trites and Heise 1996; Perry et al. 1999). Marine mammal quantitative diet information in the Caribbean was generally scarce, in particular for baleen whales which generally come to breed or calve in these waters. These migratory whales generally are not known to feed in breeding areas, but rather consume most of their food during their soujour in their subpolar or polar feeding grounds of the North Atlantic (Klinowska 1991; Clapham 2002; Jann et al. 2003).  Lockyer (1981) estimated that consumption in breeding areas is approximately 10% of feeding rates in feeding ground. Several subsequent studies (e.g. Brown and Lockyer 1984; Horwood 1990) support this estimate. This was taken into account and applied by Mohammed (2003a) in the Grenadines' ecosystem model and by Morissette et al. (submitted) in Northwest African ecosystem. Published and unpublished sources for cetacean quantitative dietary information were retrieved using several search engines (e.g. ASFA, Google scholar, Web of Science, ScienceDirect) as well as an extensive Endnote library database for marine mammals provided by Kristin Kaschner (Albert-Ludwigs-University, Freiburg, Germany, personal communication) and further updated by our team. In spite of an exhaustive literature search, very little quantitative dietary information on cetaceans in the region was found. Cetaceans, especially great whales feeding habits or diet studies are less important in the breeding areas because, as noted above, this is not where they Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  61 feed (Klinowska 1991; Perrin et al. 2002). In cases like these, we incorporated cetacean quantitative diet information from similar ecosystems where data were available. For methodological reasons, all cetacean biomass and food consumption estimates represented an annual averages, thus indirectly assuming that species remain year-round in the study area as part of the ’system’ in terms of calculated impacts (e.g. fishing, whaling and general trophic interactions). However, we represented the feeding ecology for most of the baleen whales (i.e. minke whales, humpback whales, fin whales, sei whales blue whales) more realistically by setting a high Ecopath diet proportion as ‘import‘ in the Ecopath diet consumption matrix (Christensen et al. 2005) thus assuming that the majority of the food consumption occurred outside our study area. Hence, food intake of the above-mentionned baleen whales for each prey group in the Caribbean was reduced to 10%, while the remaining 90% of the intake were considered to be taken outside of our study. Bryde’s whales were excempted from this, since the species is known to occur in Caribbean areas yearround (Mohammed et al. 2007b). Finally, our balanced trophic model, the Caribbean ecosystem, was reviewed and validated by local experts of the Caribbean region during the “whales eat fish” workshop held in Barbados on 23-25 September 2008, facilated by the project with the assistance of the Lenfest Ocean Program (http://www.lenfestocean.org/whales_fisheries.html).  Resource exploitation of the ecosystem Whaling Whaling was initiated in the Caribbean in the late 1800’s, specifically occurring in the islands of Bequia St. Vincent, the Grenadines and St. Lucia (Mahon 1993; Quimby 2000; Reeves 2002, Mohammed et al. 2007a). An approximate harvest of two to three humpback whales per season is documented for St. Vincent and the Grenadines under an “aboriginal” clause of the International Whaling Commission’s 1986 moratorium on commercial whaling (Hawley 1999; Goetz et al. 1999; Quimby 2000; Reeves 2002). Aboriginal subsistence whaling has been ongoing because it represents a valuable source of protein, and therefore whaling is considered to be an integral part of traditional practices (Hawley 1999; Quimby 2000; Perrin et al. 2002; Reeves 2002). Thus far, whales and smaller odontocetes are still harvested in some parts of the region for subsistence of indigenous peoples and for scientific research (Quimby 2000; Reeves 2002). Mohammed et al. (2007a), who recently documented catches of cetaceans in the region, states that about 140 dolphins, 13 pilot whales (Globicephala macrorhynchus) and one false killer whale (Pseudorca crassidens) have been caught annually in St. Lucia from 2000’s to 2005’s. The same study also documented catches of killer whales (5.63 tonnes/year) and shallow-diving small cetaceans (0.51 tonnes/year) from period of 2000’s to 2005’s in St. Vincent and the Grenadines. The most common small cetacean species harvested in some eastern Caribbean countries are bottlenose dolphins (Tursiops truncatus), spotted dolphins (Stenella frontalis), spinner dolphins (S. longirostris), and striped dolphins (S. coeruleoalba) (Rambally 1999; Mohammed et al. 2007a). Additionally, many small cetacean species in the Caribbean are incidentally captured in fishing gear (Mohammed et al. 2007a). Although there are some on-going whaling activities in the region (Quimby 2000; Reeves 2002, Mohammed et al. 2007a), there was no official timeseries landings found for the marine mammals catches except for killer whales and small cetaceans species. So far, only the pelagic fisheries report of Mohammed et al. (2007a) documented annual catches for killer whales and small cetacean species in Lesser Antilles countries for the period of 2001 to 2005. Fishery Local communities in the Caribbean region are highly dependent on the marine environment for their livelihood (Jeffrey 2000; Mahon and McConney 2004; Heileman 2007; Mohammed et al. Food web models and data for the Caribbean model, J.L. Melgo et al.  62 2007a). Employment and the income of locals in the region are generated in the fisheries sector mainly through commercial fisheries, recreational fishing and tourism (Opitz 1996; Mohammed et al. 2007a). Artisanal fishing (small-scale fishing) in some areas in the Caribbean is important as well, however, since near-shore resources have become depleted, fishing activities continues to expand in the open-ocean, exploiting offshores resources in the adjacent waters of the Caribbean (Mahon and McConney 2004; CLME 2007; Heileman 2007; Mohammed et al. 2007a). For instance, Mahon (1996) reported that fishery for large pelagics in Grenada has grown since the early 1980’s from 45 to 110 smal long-liners fishing one-day trips and seven short-stay longliners. In St Lucia, 45 new vessels were introduced into the large pelagics fishery between 1989 and 1992, and 82 vessels were introduced in Barbados fisheries betweeen 1979 and 1989 (Mahon 1996; Mohammed et al. 2007a). The fisheries in the Caribbean ecosystem include a diverse array of resources (Mahon 1999; Mahon and McConney 2004). The most important fisheries in the region are targeting flyingfish, dolphinfish, yellowfin tuna, albacore tuna, billfishes, swordfish, sailfish, skipjack tuna, wahoo, snappers, jacks, weakfish, reef fishes, lobsters, shrimps and conch shells (Opitz 1996; Mahon 1999; Restrepo et al. 2003; Mahon and McConney 2004; CLME 2007; Heileman 2007; Mohammed et al. 2007a). Different species of sharks (list some examples here) are also caught in the region, although most sharks are taken as bycatch (Mohammed et al. 2007a). In the southeast Caribbean (e.g. Lesser Antilles countries), fishing trips focus mainly on large pelagic fishes and flyingfish species (Mahon 1999; Mohammed et al. 2007a). For other Caribbean countries with extensive coastal shelves (e.g. Puerto Rico, the Virgin Islands), fishing is primarily targeting coastal species (reef fishes, demersal fishes crustaceans, coastal pelagics) and large pelagic fishes when they are available within the inshore areas (Mahon 1999; Mohammed et al. 2007a). The most common fishing gear used includes gillnets, trolling, longline, beach seines, purse seines, hooks and dredges (Opitz 1996; Cummings and Matos-Caraballo 2007; Mohammed et al. 2007a). In addition, large pelagics are targeted using smaller and larger vessels employing gillnets, troll nets and long-lines (Mahon and McConney 2004; Mohammed et al. 2007a). The timeseries catch data (1987-2005) used in the developd model of the Caribbean ecosystem were obtained both from the report of Lesser Antilles Pelagic Ecosystem (LAPE) project (Mohammed et al. 2007a) and from the Sea Around Us database (unpublished database supplemented by Reg Watson of the Sea Around Us Project and www.seaaroundus.org) (Tables 2-4). The LAPE fisheries report contained reported 2001-2005 landings within the Lesser Antilles countries’ EEZ (Antigua and Barbuda, Barbados, Dominica, Grenada and the Grenadines, Guadeloupe, Martinique, St Kitts and Nevis, St Lucia, St Vincent, and Trinidad and Tobango) compiled from a number of different sources. In order to cover our entire study area, which encompassed but did not exactly match the LAPE project area, and to make the best use of data on timeseries of catches in the area for model analysis, we included catches from the Sea Around Us Project (SAUP) covering the earlier period of 1987-2004 (unpublished database supplemented by Reg Watson of the Sea Around Us Project and www.seaaroundus.org,). The latter also includes catches of foreign fleets fishing in the EEZs of Caribbean nations starting from 1987 to 2004. It should be noted that SAUP catch data for specific areas is derived using a complex set of databases to attribute specific catches reported to FAO back to their geographic origin (Watson et al. 2004). Local fleets were defined as all fisheries operated by countries encompassed by our study area, regardless of whether or not catches were taken within each country’s own EEZ waters or neighboring waters. All other fishing countries were defined as foreign fleets. Local fleets included fisheries from Anguilla, Antigua and Barbuda, Barbados, British Virgin Islands, Dominica, Dominican Republic, Grenada and the Grenadines, Guadeloupe, Martinique, Montserrat, Puerto Rico, St Kitts and Nevis, St Lucia, St Vincent, and Trinidad and Tobango, and US Virgin Island. The foreign fleets fishing the region were primarily from USA and Puerto Rico, while only a very small percentage of catches in the area were made by fleets from Venezuela, Netherland Antilles and Colombia. The comparison of both databases showed that the small and schooling pelagic group were harvested the most by both local (SAUP and LAPE databases) and foreign fleets (SAUP). Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  63 In order to combine fisheries catch data sets from LAPE and SAUP, we calculated the catch densities for each trophic group in the model. For example, local catches from 1987 to 2000 for each trophic group were taken from SAUP. These catches were then divided by the total Caribbean area defined for SAUP database extraction (7778039.50 km-2). Local catches for the 2001-2005 periods were obtained from both LAPE database (for countries of the Lesser Antilles) and SAUP database for all remaining countries (local in our study area but not covered by LAPE). We divided their catch data for each trophic group by its total areas i.e. catches from Sea Around Us Project were divided by 7778039.50 km-2 and catches from LAPE fisheries report were divided by 610,000.00 km-2. These catch densities were then summed by year and trophic groups resulting the total local catches for the 2001-2005 periods in our study area. Catch densities of foreign fleets were only available from the SAUP database, thus, we used these data sets for the period of 1987 to 2004 and calculated the catch densities for trophic groups accordingly. Generally, local fleets (~ 87%) contributed largely to the overall catches in the Caribbean in comparison to the foreign fleets (~13%). The majority of these catches were comprised of the small and schooling pelagics, averaging 0.032 t*yr*km-2 for local fleets, and 0.002 t*yr*km-2 for foreign fleets in the Caribbean (Tables 5 and 6). According to Heileman-Manickchand (1992), huge quantities of small schooling pelagics (e.g. sardines, herrings) are caught in the region for bait in commercial and recreational fishing in the area, and some are utilized for canned and smoked products. Catches of crustaceans and benthos, scombrids, reef fishes, coastal predators and small tunas were also important in terms of tonnage for local fleets of the Caribbean region (Table 5), while for foreign fleets; these species were moderately important landings (Table 6). The local catches for cetaceans and seaturtles were low in the region (Table 5), and not recorded for foreign fleets (Table 6). Fisheries catch trends of all trophic groups caught by local fleets in the area fluctuated considerably between 1987’s to 2005’s (Figure 2). The fluctuation hit its peak around 2003. After that peak, catch densities fell to a record low in 2005 (Figure 2). In contrast, the catch density trends for foreign fleets showed distinct fluctuation with two peaks around 1994 and 1997. They also had two major low points on 1996 and 1999. In the subsequent years catches of both local and foreign fisheries have decreased substantially.   Food web models and data for the Caribbean model, J.L. Melgo et al.  64 Table 2. Local fleets’ catches (tonnes) by trophic group in the Caribbean region. Catch data were derived from Sea Around Us Project supplemented by input provided by Reg Watson and www.seaaroundus.org. 13 14 15 16 17 18 19 20 21 22 23 24 25 Year Large tunas and billfishes Small tunas Dolphin- fish Flying- fish Other offshore predators Pelagic sharks Coastal sharks and rays Scombrids Small and schooling pelagics Reef fishes Coastal predators Cepha- lopods Crustaceans and benthos 1987 585.04 778.60 656.33 91.57 0.02 90.97 285.32 3435.73 17499.41 4696.55 2038.29 68.63 3097.91 1988 657.03 803.20 657.96 121.37 0.02 58.19 217.48 4014.71 11902.86 3580.05 2072.09 72.16 4239.27 1989 601.64 1218.37 749.14 28.04 0.03 55.74 256.94 4130.20 13886.08 4050.54 2368.14 88.70 7044.68 1990 691.44 1431.29 832.50 33.31 0.03 52.63 219.38 4092.50 13251.95 3828.67 2370.38 77.05 6946.97 1991 831.68 1311.43 873.63 85.53 0.03 80.24 265.54 3578.67 14443.72 4273.78 2181.19 81.06 7203.85 1992 859.04 1145.20 905.11 157.46 0.03 76.55 158.12 4318.24 12120.32 2982.40 2452.37 98.03 5921.84 1993 1022.77 1093.26 901.28 140.44 0.04 81.21 121.63 4148.43 11957.31 3453.40 2314.77 43.35 5894.89 1994 1015.47 1300.84 1160.77 83.87 0.05 69.03 155.78 3583.23 11058.42 5854.14 3412.48 37.57 7988.82 1995 967.07 1530.66 1089.53 99.56 0.05 73.54 198.93 4548.46 11509.30 5706.56 2822.23 61.26 5927.83 1996 890.67 1220.98 1238.78 133.23 0.06 57.72 167.25 4725.64 12455.30 2448.73 3121.59 69.48 4321.04 1997 724.62 1114.64 1248.20 77.73 0.06 60.59 204.10 4127.40 13776.04 3449.33 3054.46 54.53 5102.19 1998 548.37 1175.56 1001.51 182.08 0.04 65.77 174.86 3465.81 12163.74 1551.59 2329.85 141.44 6571.06 1999 423.74 867.24 1291.39 115.00 0.04 72.61 241.34 2741.91 13984.83 2246.66 1699.98 76.91 5471.79 2000 964.50 937.74 1305.87 143.41 0.03 61.77 220.42 3187.50 14159.13 2245.15 2611.97 176.97 7943.62 2001 323.16 344.26 141.55   4.33 92.28 504.20 4622.35 1989.00 2010.32 112.95 4653.39 2002 343.57 423.28 339.45   2.35 12.26 642.22 2466.69 2743.37 2679.67 141.83 6467.29 2003 803.17 426.06 119.81   1.94 89.93 629.27 6556.28 2541.70 3744.34 137.91 4680.39 2004 333.62 95.13 134.02    39.64 600.96 4893.51 1804.28 2388.27 89.67 4442.54  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  65 Table 3. Local fleets’ catches (tonnes) by trophic group in Lesser Antilles. Catch data were derived from LAPE fisheries report by Mohammed et al. (2007a) . 7 10 12 13 14 15 16 18 19 20 21 22 23 24 Year Killer whales Small ceta- ceans Sea turtles Large tunas and billfishes Small tunas Dolphin -fish Flying- fish Pelagic sharks Coastal sharks and rays Scom- brids Small and schooling pelagics Reef fishes Coastal predators Cepha- lopods 2001 7.87 13.54 7.00 1965.04 2752.12 2457.15 2085.67 102.77 893.90 3750.73 7688.02 226.44 417.01 5.00 2002 7.63 10.54 12.44 1553.25 3705.20 2396.07 1842.93 72.66 1210.22 3885.24 7475.29 232.08 529.36 3.00 2003 6.37 11.19 13.19 1955.43 2871.73 2082.34 2021.02 116.14 810.71 2973.89 7325.05 243.31 363.13 2004 4.73 9.38 11.32 1805.02 2878.99 2173.95 1210.70 107.38 990.90 2935.74 7205.89 271.16 347.59 2005 3.14 8.42 33.87 2189.97 2738.80 1875.68 1192.85 80.42 460.19 3399.43 6660.05 188.53 407.78  Food web models and data for the Caribbean model, J.L. Melgo et al.  66 Table 4. Foreign fleets’ catches (tonnes) by trophic group in the Caribbean region. Catch data were derived from Sea Around Us Project supplemented by input provided by Reg Watson and www.seaaroundus.org. 13 14 15 17 18 19 20 21 22 23 24 25 Year Large tunas and billfishes Small tunas Dolphin- fish Other offshore predators Pelagic sharks Coastal sharks and rays Scom- brids Small and schooling pelagics Reef fishes Coastal predators Cepha- lopods Crustaceans and benthos 1987 456.29 23.48 7.52 140.74 7.85 90.21 607.06 2131.93 528.26 942.73 14.76 815.26 1988 400.86 34.73 13.38 135.00 23.57 177.54 616.97 2292.80 739.60 1096.68 10.47 876.53 1989 317.18 26.67 26.25 104.00 30.03 225.50 560.26 2300.13 903.05 1109.74 1.41 825.25 1990 165.53 33.47 30.50 118.35 26.77 188.53 571.57 2747.85 804.29 995.45 1.48 855.70 1991 158.59 39.51 40.14 111.94 21.54 144.48 579.12 2069.55 794.24 1020.75 16.63 1072.12 1992 308.97 71.63 21.49 92.81 32.04 159.15 601.10 1967.13 682.04 902.55 1.81 815.23 1993 135.48 53.91 21.07 86.59 29.47 123.31 740.98 2402.84 875.26 1173.11 298.09 2062.55 1994 116.88 34.38 26.33 131.65 52.35 535.05 700.53 1983.81 1130.24 3157.13 8.30 1337.35 1995 357.93 21.60 41.62 41.54 40.36 116.24 614.77 1928.09 723.24 891.40 12.92 1352.32 1996 97.90 18.35 31.39 10.24 13.36 46.59 186.35 1156.23 594.69 644.37 0.53 1115.07 1997 184.98 44.66 32.41 30.81 35.67 114.98 723.37 2411.06 576.01 862.41 9.02 1191.82 1998 114.96 36.83 13.68 31.74 39.05 115.26 690.34 1590.97 635.49 792.81 17.24 999.40 1999 95.88 27.81 23.57 21.92 25.37 69.39 441.22 504.63 157.47 509.66 7.87 896.25 2000 117.41 28.98 21.15 20.52 115.38 61.68 663.11 1163.59 742.75 757.01 4.71 820.54 2001 52.65 34.87 17.46 20.11 107.29 56.36 674.99 900.27 724.96 778.43 9.50 621.19 2002 64.21 38.47 16.52 17.87 37.14 74.95 640.93 1314.33 750.15 751.36 14.65 692.70 2003 66.44 55.77 16.18 17.60 43.36 81.49 720.55 995.19 755.43 817.45 9.98 1376.03 2004 64.89 20.17 16.14 32.49 64.27 192.47 787.79 919.15 680.26 772.56 3.04 709.48  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  67 Table 5. Local fleets’ catch densities (tonnes*yr*km-2) of each trophic group in the Caribbean. Data were derived from Sea Around Us Project (supplemented by input providedby Reg Watson of SAUP and www.seaaroundus.org) and LAPE fisheries report (Mohammed et al. 2007a) 7 10 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Year Killer whales Small ceta- ceans Sea turtles Large tunas and bill- fishes Small tunas Dolphin -fish Flying- fish Other offsh. preda- tors Pelagic sharks Coastal sharks and rays Scom- brids Small and school. pelagics Reef fishes Coastal pred. Cepha- lopods Crusta- ceans and benthos 1987    7.519E-04 1.001E-03 8.436E-04 1.177E-04 2.320E-08 1.169E-04 3.667E-04 4.416E-03 2.249E-02 6.036E-03 2.620E-03 8.821E-05 3.982E-03 1988    8.445E-04 1.032E-03 8.457E-04 1.560E-04 2.812E-08 7.480E-05 2.795E-04 5.160E-03 1.530E-02 4.601E-03 2.663E-03 9.275E-05 5.449E-03 1989    7.733E-04 1.566E-03 9.629E-04 3.605E-05 3.232E-08 7.164E-05 3.302E-04 5.308E-03 1.785E-02 5.206E-03 3.044E-03 1.140E-04 9.054E-03 1990    8.887E-04 1.840E-03 1.070E-03 4.282E-05 3.562E-08 6.764E-05 2.820E-04 5.260E-03 1.703E-02 4.921E-03 3.047E-03 9.903E-05 8.929E-03 1991    1.069E-03 1.686E-03 1.123E-03 1.099E-04 4.011E-08 1.031E-04 3.413E-04 4.600E-03 1.856E-02 5.493E-03 2.803E-03 1.042E-04 9.259E-03 1992    1.104E-03 1.472E-03 1.163E-03 2.024E-04 4.442E-08 9.839E-05 2.032E-04 5.550E-03 1.558E-02 3.833E-03 3.152E-03 1.260E-04 7.611E-03 1993    1.315E-03 1.405E-03 1.158E-03 1.805E-04 4.672E-08 1.044E-04 1.563E-04 5.332E-03 1.537E-02 4.439E-03 2.975E-03 5.572E-05 7.577E-03 1994    1.305E-03 1.672E-03 1.492E-03 1.078E-04 6.273E-08 8.872E-05 2.002E-04 4.605E-03 1.421E-02 7.524E-03 4.386E-03 4.828E-05 1.027E-02 1995    1.243E-03 1.967E-03 1.400E-03 1.280E-04 7.004E-08 9.452E-05 2.557E-04 5.846E-03 1.479E-02 7.335E-03 3.627E-03 7.874E-05 7.619E-03 1996    1.145E-03 1.569E-03 1.592E-03 1.712E-04 7.514E-08 7.419E-05 2.150E-04 6.074E-03 1.601E-02 3.147E-03 4.012E-03 8.930E-05 5.554E-03 1997    9.313E-04 1.433E-03 1.604E-03 9.990E-05 7.190E-08 7.788E-05 2.623E-04 5.305E-03 1.771E-02 4.433E-03 3.926E-03 7.009E-05 6.558E-03 1998    7.048E-04 1.511E-03 1.287E-03 2.340E-04 5.428E-08 8.453E-05 2.247E-04 4.455E-03 1.563E-02 1.994E-03 2.995E-03 1.818E-04 8.446E-03 1999    5.446E-04 1.115E-03 1.660E-03 1.478E-04 4.896E-08 9.332E-05 3.102E-04 3.524E-03 1.797E-02 2.888E-03 2.185E-03 9.885E-05 7.033E-03 2000    1.240E-03 1.205E-03 1.678E-03 1.843E-04 4.487E-08 7.939E-05 2.833E-04 4.097E-03 1.820E-02 2.886E-03 3.357E-03 2.275E-04 1.021E-02 2001 1.290E-05 2.219E-05 1.148E-05 3.637E-03 4.954E-03 4.210E-03 3.419E-03  1.740E-04 1.584E-03 6.797E-03 1.854E-02 2.928E-03 3.267E-03 1.534E-04 5.981E-03 2002 1.252E-05 1.728E-05 2.039E-05 2.988E-03 6.618E-03 4.364E-03 3.021E-03  1.221E-04 2.000E-03 7.195E-03 1.542E-02 3.906E-03 4.312E-03 1.872E-04 8.312E-03 2003 1.043E-05 1.834E-05 2.162E-05 4.238E-03 5.255E-03 3.568E-03 3.313E-03  1.929E-04 1.445E-03 5.684E-03 2.043E-02 3.666E-03 5.408E-03 1.772E-04 6.016E-03 2004 7.761E-06 1.537E-05 1.855E-05 3.388E-03 4.842E-03 3.736E-03 1.985E-03  1.760E-04 1.675E-03 5.585E-03 1.810E-02 2.764E-03 3.639E-03 1.153E-04 5.710E-03 2005 5.141E-06 1.380E-05 5.553E-05 3.590E-03 4.490E-03 3.075E-03 1.955E-03  1.318E-04 7.544E-04 5.573E-03 1.092E-02 3.091E-04 6.685E-04  Food web models and data for the Caribbean model, J.L. Melgo et al.  68 Table 6. Foreign fleets’ catch densities (tonnes*yr*km-2) of each trophic group in the Caribbean. Data were derived from Sea Around Us supplemented by input provided from Reg Watson and www.seaaroundus.org. 13 14 15 17 18 19 20 21 22 23 24 25 Year Large tunas and billfishes Small tunas Dolphin- fish Other offshore predators Pelagic sharks Coastal sharks and rays Scom- brids Small and schooling pelagics Reef fishes Coastal predators Cepha- lopods Crustaceans and benthos 1987 0.000586 0.000030 0.000010 0.000181 0.00001 0.00012 0.00078 0.00274 0.000679 0.001212 1.89732E-05 0.00104784 1988 0.000515 0.000045 0.000017 0.000174 0.00003 0.00023 0.000793 0.002947 0.000951 0.00141 1.34629E-05 0.00112659 1989 0.000408 0.000034 0.000034 0.000134 0.00004 0.00029 0.00072 0.002956 0.001161 0.001426 1.81754E-06 0.00106068 1990 0.000213 0.000043 0.000039 0.000152 0.00003 0.00024 0.000735 0.003532 0.001034 0.001279 1.90863E-06 0.00109982 1991 0.000204 0.000051 0.000052 0.000144 0.00003 0.00019 0.000744 0.00266 0.001021 0.001312 2.13798E-05 0.00137798 1992 0.000397 0.000092 0.000028 0.000119 0.00004 0.00020 0.000773 0.002528 0.000877 0.00116 2.33207E-06 0.00104781 1993 0.000174 0.000069 0.000027 0.000111 0.00004 0.00016 0.000952 0.003088 0.001125 0.001508 0.000383128 0.00265096 1994 0.00015 0.000044 0.000034 0.000169 0.00007 0.00069 0.0009 0.00255 0.001453 0.004058 1.0674E-05 0.00171887 1995 0.00046 0.000028 0.000053 5.34E-05 0.00005 0.00015 0.00079 0.002478 0.00093 0.001146 1.66035E-05 0.00173811 1996 0.000126 0.000024 0.000040 1.32E-05 0.00002 0.00006 0.00024 0.001486 0.000764 0.000828 6.83352E-07 0.00143318 1997 0.000238 0.000057 0.000042 3.96E-05 0.00005 0.00015 0.00093 0.003099 0.00074 0.001108 1.15883E-05 0.00153183 1998 0.000148 0.000047 0.000018 4.08E-05 0.00005 0.00015 0.000887 0.002045 0.000817 0.001019 2.21631E-05 0.00128451 1999 0.000123 0.000036 0.000030 2.82E-05 0.00003 0.00009 0.000567 0.000649 0.000202 0.000655 1.01135E-05 0.00115193 2000 0.000151 0.000037 0.000027 2.64E-05 0.00015 0.00008 0.000852 0.001496 0.000955 0.000973 6.04873E-06 0.00105463 2001 6.77E-05 0.000045 0.000022 2.58E-05 0.00014 0.00007 0.000868 0.001157 0.000932 0.001001 1.22057E-05 0.0007984 2002 8.25E-05 0.000049 0.000021 2.3E-05 0.00005 0.00010 0.000824 0.001689 0.000964 0.000966 1.88268E-05 0.00089032 2003 8.54E-05 0.000072 0.000021 2.26E-05 0.00006 0.00010 0.000926 0.001279 0.000971 0.001051 1.2821E-05 0.00176859 2004 8.34E-05 0.000026 0.000021 4.18E-05 0.00008 0.00025 0.001013 0.001181 0.000874 0.000993 3.91262E-06 0.00091188   Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  69   0.000 0.010 0.020 0.030 0.040 0.050 0.060 0.070 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 1 9 9 7 1 9 9 8 1 9 9 9 2 0 0 0 2 0 0 1 2 0 0 2 2 0 0 3 2 0 0 4 2 0 0 5 0.000 0.002 0.004 0.006 0.008 0.010 0.012 0.014 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 Killer w hales Small cetaceans Seaturtles Large tunas and billf ishes Small tunas Dolphinfish Flyingfish Other offshore predators Pelagic sharks Coastal sharks and rays Scombrids Small and schooling pelagics Reef f ishes Coastal predators Cephalopods Crustaceans and benthos  Figure 2. Trophic groups catch densities for local fleets (1987-2005) and foreign fleets (1987-2004) in the Caribbean region. Data derived from SAUP (supplemented by data provided by Reg Watson and www.seaaroundus.org) and LAPE fisheries report (Mohammed et al. 2007a)  Local fleets Foreign fleets  Local fleets C at ch  d en si ty  (t on ne s* yr *k m -2 ) Food web models and data for the Caribbean model, J.L. Melgo et al.  70 BALANCED ECOSYSTEM MODEL FOR THE CARIBBEAN REGION Groups 1 - 10. Marine mammals Ten groups of marine mammals were added to the original model for our study. Given the lack of local, long-term dedicated surveys to provide reliable cetacean abundance estimates, density estimates had to be derived from a global database (Kaschner 2004). However, comparison with other densities from surveys conducted in similar habitats are ground-truthing these estimates (Table 7). These biomasses, P/B and diet are described below for each group, followed by a section on how we calculated food consumption (Q/B) for these ten groups. As mentioned above, no official timerseries records of annual catches for marine mammals, excluding killer whales and small cetacean species, were found in the region.  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  71 Table 7: Comparison of predicted cetacean densities in Caribbean waters based on global model developed by Kaschner et al (2006) and Kaschner (2004) and observed minimum and maximum densities in similar habitats (subtropical & tropical waters). A = aerial surveys, S = ship based surveys.  Density estimates that are corrected for animals missed on the track-line are indicated in the G(0) corrected column. All other observed estimates might represent underestimations. Common Name Estimated density [animals / 1000 km2] Observed density [animals / 1000 km2] CV G(0) corrected Geographic area Survey years Survey type Source Blue whale 0.02 0.07 0.24 no Eastern Tropical Pacific 1986-1990 S Wade and Gerrodette 1993 Blue whale 0.02 4.96 0.13 yes NE Pacific, California inshore 1991-1996 S Calambokidis and Barlow 2004 Bryde's whale 0.28 0.67 0.20 no Eastern Tropical Pacific 1986-1990 S Wade and Gerrodette 1993 Bryde's whale 0.28 0.04 0.85 no NW Atlantic, northern Gulf of Mexico (GulfCet I survey) 1991-1994 S Davis and Fargion 1996 Fin whale 0.22 1.85 0.48 no NW Atlantic, Virginia Capes 2002 S Garrison et al. 2003 Fin whale 0.22 0.07 0.72 yes NE Pacific, Hawaiin waters 2002 S Barlow 2003a Humpback whale 6.50 0.03 0.37 yes NE Pacific, California offshore 1991-1996 S Calambokidis and Barlow 2004 Humpback whale 6.50 112.32 0.27 no SW Indian Ocean, Madagaskar (southern block) 1994 S Best et al. 1996 Minke whale 0.73 0.93 0.51 yes NE Pacific, west coast US 1996 S Barlow 2003b Minke whale 0.73 0.03 1.29 no NW Atlantic, US east coast, south of Maryland 1998 S Mullin and Fulling 2003 Sei whale 0.05 0.10 0.73 yes NE Pacific, west coast US 1996 S Barlow 2003b Sei whale 0.05 0.03 1.01 yes NE Pacific, west coast US 2001 S Barlow 2003b Killer whale 0.06 0.14 0.98 yes NE Pacific, Hawaiin waters 2002 S Barlow 2006 Killer whale 0.06 0.79 0.48 no NW Atlantic, northern Gulf of Mexico (GulfCet I survey) 1991-1994 S Davis and Fargion, 1996 Sperm whale 1.03 0.85 0.57 no NW Atlantic, northern Gulf of Mexico (GulfCet I EPA survey) 1996-1997 S Davis et al. 2000 Sperm whale 1.03 3.80 0.23 no NW Atlantic, northern Gulf of Mexico (SEFSC) 1996-2001 S Mullin and Fulling 2004  Food web models and data for the Caribbean model, J.L. Melgo et al.  72 1. Minke whales Although this represents probably the southern most extent of this species distributional range, minke whales are seasonally found in the northwestern Caribbean during the breeding period (Klinowska 1991; Perrin et al. 2002). Their annual abundance in the region was estimated to be 1,352 individuals, translating to a density of 0.73 individuals per 1000 km2 and representing a total annual biomass of 8,882 tonnes, or biomass density of 0.0045 t*km-2 annually based on a global model of cetacean densites developed by Kaschner (2004). The predicted densities (individuals per 1000 km2) were within the range of observed densities reported from dedicated marine mammals surveys conducted in other areas with similar habitats (Table 7). The reported general annual mortality of minke whales ranges from 0.09 to 0.10 yr-1 in Northeast Atlantic (Evans 1998). We assumed that this mortality rate of minke whales would also apply in the Caribbean region. Since the annual mortality rate is equal to the production to biomass (P/B) ratio (Allen 1971), an annual mortality rate of 0.099 yr-1 was used as the P/B input for minke whales in the present model. This value was also used for minke whales in Alaska (Guénette et al. 2006) and in Northwest African model (Morissette et al. submitted) There was no appropriate quantitative diet information available for minke whales in the Caribbean region. Hence, we used the available data for this species from North Atlantic (Table 8). Minke whale diet information from the different published studies was redistributed to match our pre-defined trophic groups. These were then weighted, based on the proportion of prey wet weight over the overall prey weights in whale’s stomach. Most of the minke whales’ diet (49.0%) was composed of small and schooling pelagics followed by zooplankton (25.9%), coastal predators (12.1%), other offshores pelagics (9.8%) and scombrids (3.2%) (Table 7). Table 8.  Available information on the diet composition of minke whales in the Caribbean. The average diet was used in the Ecopath model for the Caribbean ecosystem. 17 20 21 23 26 Source Areas  Other offshore predators Scombrids Small and schooling pelagics Coastal predators Zooplank- ton Total Lydersen et al. 1991 Norway 0.084  0.916   1.000 Nørdoy and Blix 1992 Northeastern Atlantic 0.078  0.501  0.421 1.000 Haug et al. 1995 Norwegian waters 0.126  0.708 0.084 0.082 1.000 Haug et al. 1996 Northeast Atlantic 0.343  0.388 0.026 0.244 1.000 North Sea 0.029 0.093 0.011 0.867  1.000 Olsen and Holst 2001 Norwegian Sea   1.000   1.000 Sivertsen et al. 2006 Barent Sea 0.090  0.433  0.477 1.000 Smout and Lindstrom 2007 Norwegian Sea   0.261  0.739 1.000 Bear Island 0.096  0.504  0.400 1.000 North Sea 0.087 0.317  0.597  1.000 Norwegian Sea 0.036  0.916  0.049 1.000 Southern Barent Sea 0.247  0.713 0.004 0.036 1.000 Windsland et al. 2007 Spitsbergen 0.062  0.016  0.922 1.000 Average Minke whale diet 0.098 0.032 0.490 0.121 0.259 1.000  2. Fin whales Fin whales, breeding in the Caribbean ecosystem are typically migrants from the North Atlantic (Klinowska 1991; Perrin et al. 2002), although, like the minke whales, this species is probably only found rarely as for south as our study area. During the recent cetaceans sighting survey by NOAA Gordon Gunter, one fin whale was sighted in the Lesser Antilles waters (Mohammed et al. 2007b). Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  73 According to the quantitative estimates by Kaschner (2004), approximately 139 fin whales occur annually in the Caribbean, representing a density of 0.22 22 individuals per 1000 km2 and an annual biomass of 7742 tonnes or average annual biomass density of 0.0062 t*km-2. The predicted density (individuals per 1000 km2) was in the same range as observed densities reported from dedicated marine mammals surveys conducted in other areas with similar habitats (Table 7). Here again, we used the total annual mortality rate as the P/B input in the model, following Allen (1971). Different studies reported the natural mortality rates ranging from 0.4 yr-1 to 0.5 yr-1 for adult fin whales (Clarke 1982; de la Mare 1985; Perry et al. 1999). Perry et al. (1999) stated that this rate may be higher if immature whales of both sexes were included. On the other hand, Heymans (2005) used P/B ratio of 0.099 yr-1 for fin whales in Aleutian Islands and in the Gulf of Alaska. We used this value in the present model and assumed it to be constant as well in the Caribbean. The same P/B ratio value for fin whales was used as well by Morissette et al. (submitted) in Northwest Africa model. In the absence of diet information for fin whales in the Caribbean, we used diet information for this species that was available from Icelandic waters Sigurjónsson and Víkingsson (1997). Based on that, it was thus assumed that fin whales eat 98.4% zooplankton, 1.4% coastal pelagics and 0.2% coastal predators (Table 9). Note that a small proportion of ‘mixture remains’ identified by Sigurjónsson and Vikíngsson (1997) were not included herein because this represents unknown prey type. Table 9. Diet composition of fin whales used for Ecopath modelling in the Caribbean ecosystem. 21 23 26 Source Areas  Small and schooling pelagics Coastal predators Zooplankton Total Sigurjónsson and Víkingsson 1997 Icelandic and adjacent waters 0.014 0.002 0.984 1.000  3. Humpback whales Known breeding areas of the north Atlantic humpback whale population include a number of areas in and around the Lesser Antilles, Dominican Republic, Puerto Rico, the Virgin Islands and the South Grenadines (Katona and Beard 1990; Klinowska 1991; Debrot et al. 1998; Perry et al. 1999; Reeves et al. 2001; Romero et al. 2001; Swartz et al. 2003). This whale species calves in the Caribbean ecosystem from January to May (Debrot et al. 1998; Reeves et al. 2001; Romero et al. 2001; Swartz et al. 2003). All individuals from the total North Atlantic population of this species [about 10,750 whales estimated by Stevick et al. (2003), translating to 6.50 individuals per 1000 km2, and representing 326,888 tonnes in biomass] are thought to migrate to the Caribbean for breeding at some stage during the winter. However, since the species occurrence in the area is restricted to about five months (Reeves et al. 2001), we calculated annual biomass density proportionally, thus obtaining an estimate of 0.0696 t*km-2. The latter value was used in the present model as the overall annual biomass density of humpback whales in the Caribbean ecosystem, and is similar to the range of observed densities reported from dedicated marine mammals surveys conducted in other areas with similar habitats (Table 7). Similar to other baleen whale species, P/B ratio of 0.099 used by Heymans (2005) in Aleutian Islands and in the Gulf of Alaska was used for humpback whale in our model. This value was assumed to be similar in the Caribbean region as well. Romero et al. (2001) stated that humpback whales do not feed in the Caribbean region, and no diet information of this species could be found for the region. We, therefore, used two available quantitative diet studies providing diet information in wet weight available for humpback whales: one by Witteveen et al. (2006) in Kodiak, Alaska, and the other one by Mitchell (1973) in Canadian waters which was then used by Sigurjónsson and Vikíngsson (1997) in Icelandic waters. We incorporated the “fish” and “krill” ratio of 60:40 from Mitchell (1973) obtained in the Canadian waters to quantify the trophic groups used in the present model that were based on the more recent and detailed humpback fish prey list found in Witteveen et al. (2006). Thus, the diet of humpback whales in the Caribbean model was assumed to be 40.0% zooplankton, 34.3% small and schooling pelagics, 8.6% other offshore pelagics  and 1.7% coastal predators (Table 10). Food web models and data for the Caribbean model, J.L. Melgo et al.  74 Table 10. Diet composition of humpback whales used for Ecopath model in the Caribbean. 17 21 23 26 Source Areas Other offshore predators Small and schooling pelagics Coastal predators Zooplankton Total Sigurjónsson and Víkingsson 1997 Icelandic and adjacent waters 0.086 0.343 0.171 0.400 1.000   4. Bryde’s whales Bryde’s whales are the only species of baleen whales known to occur year-round in the Caribbean region, particularly in the southeastern Caribbean (Klinowska 1991; Debrot et al. 1998; Romero et al. 2001). Bryde’s whales are usually found off Venezuela, and often associated with schools of Sardinella anchovia (Romero et al. 2001). The annual abundance of this species in our study area was estimated to be 524 whales representing a density of 0.28 individuals per 1000 km2 and a total biomass 8,463 tonnes or an annual biomass density of 0.0043 t*km-2 based on the global density estimates by Kaschner (2004). Predicted and observed densities reported from dedicated marine mammals surveys conducted in other areas with similar habitats were similar in terms of magnitude (Table 7). In general, P/B ratios for baleen whales used in the Ecopath models in other ecosystems ranged from 0.02 yr-1 to 0.099 yr-1 (Pauly et al. 1996; Trites and Heise 1996; Zeller and Freire 2001; Okey et al. 2004; Guénette et al. 2006). Since there was no P/B ratio available, specifically for Bryde’s whales; hence, we took an average value of 0.05 yr-1 based on all reported other baleen whales P/B ranges. We used the latter value in the present model representing Bryde’s whales in the Caribbean ecosystem. Bryde’s whale diet information was taken from the Gulf of California, Mexico (Tershy 1992) and Lesser Antilles pelagic ecosystem (Heileman et al. 2007). This species primariy feeds on small and schooling pelagics (53.6.5%), zooplankton (24.5%) and other offshores pelagics (14.9%). They also feed occasionally on scombrids fishes (7.0%) (Tershy 1992; Heileman et al. 2007) (Table 11). Table 11. Available information on Bryde’s whales’ diet composition used in the Ecopath model for the Caribbean ecosystem. 17 20 21 26 Source Areas Other offshore predators Scombrids Small and schooling pelagics Zooplankton Total Tershy 1992 Gulf of California, Mexico   0.886 0.114 1.000 Heileman et al. 2007 Lesser Antilles 0.297 0.141 0.187 0.375 1.000 Average Bryde’s whale diet 0.149 0.070 0.536 0.245 1.000  5. Sei whales The sei whales are known to regularly occur around Caribbean islands area during its breeding period (Klinowska 1991; Perrin 2002). We used an annual abundance estimate of 99 whales occurring in the region, representing a density of 005 individuals per 1000 km2 for a total biomass of 1,667 tonness or an annual biomass density of 0.0009 t*km-2 (Kaschner 2004). The predicted densities (individuals per 1000 km2) were similar in terms of magnitude as observed densities reported from dedicated marine mammals surveys conducted in other areas with similar habitats (Table 7). The P/B for sei whales in the Caribbean ecosystem was assumed to be similar to what was used for this species in the Alaska Gyre ecosystem (Pauly et al. 1996), which was 0.0200 yr-1. In the absence of any diet information for sei whale in the Caribbean region, we incorporated the sei whale diet information from Sigurjónsson and Víkingsson (1997) in Icelandic waters. Sei whales’ diet Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  75 was composed of zooplankton (98.0%), other offshore predators (0.67%), small and schooling pelagics (0.67%) and coastal predators (0.67%) (Table 12). Table 12. Available information on the dietary composition of sei whale used in the Ecopath model for the Caribbean. 17 21 23 26 Source Areas Other offshore predators Small and schooling pelagic Coastal predators Zooplankton Total Sigurjónsson and Víkingsson 1997 Icelandic and adjacent waters 0.0067 0.0067 0.0067 0.980 1.000  6. Blue whales Similar to the other large whales, the blue whales, Balaenoptera musculus, are also found in the Caribbean during the breeding period (Klinowska 1991). Based on the estimated global densities provided by Kaschner (2004), the annual abundance of blue whales in the study area was approximately 47 whales representing a density of 0.02 individuals per 1000 km2 for a total biomass 4,852 tonnes and an annual biomass density of 0.0025 t* km-2. The predicted density (individuals per 1000 km2) was similar in terms of magnitude as observed densities reported from dedicated marine mammals surveys conducted in other areas with similar habitats and showed in Table 7. We used an average P/B value of 0.06 yr-1 for baleen whales in the Caribbean, based on that of several baleen whales groups from other Ecopath models (Pauly et al. 1996; Okey et al. 2004; Guénette et al. 2006). This P/B ratio was also used for the same species in Northwest Africa (Morissette et al. submitted). Quantitative diet information for blue whales was lacking in the Caribbean region. Thus, we used the information used by Sigurjónsson and Víkingsson (1997) in Icelandic waters to calculate the diet of baleen whales. The blue whales in North Atlantic mainly consumed zooplankton (i.e. krill) (Table 13). This finding was affirmed by several experts (e.g. Hjort and Ruud 1929; Klinowska 1991; Christensen et al. 1992; Tershy 1992; Pauly et al. 1998b; Hewitt and Lipsky 2002; Sears 2002) that have described blue whales ecology and distribution. Table 13. Available information on the dietary composition of blue whale used in the Ecopath model for the Caribbean. 26 Source Areas Zooplankton Total Sigurjónsson and Víkingsson 1997 Icelandic and adjacent waters 1.000 1.000  7. Sperm whales Based on life history similarities, we combined the following three species into a single trophic group: Kogia breviceps (pygmy sperm whales), Kogia simus (drawf sperm whales) and Physeter macrocephalus (sperm whales). These species are frequent in the Caribbean ecosystem from October to March (Debrot et al. 1998; Mohammed et al. 2007b) and rarely present during the summer months (Northridge 1984). The sperm whales are distributed in the deeper basins of the Caribbean Sea and the Gulf of Mexico (Klinowska 1991; Perry et al. 1999). Their annual abundance in the Caribbean was 2,154 whales, for a density of 1.03 individuals per 1000 km2, and an annual biomass of 36,904 tonnes or an annual biomass density of 0.0188 t*km-2 (Kaschner 2004). This estimated density (in individuals per 1000 km2) is similar in terms of magnitude as observed densities reported from dedicated marine mammals surveys conducted in other areas with similar habitats (Table 7). The annual total mortality rate for sperm whales species in the Caribbean region was 0.05 yr-1 (Perry et al. 1999) and was used as the P/B inputs in the present model after Allen (1971). This value falls within the reported mortality rates for different sexes and life stages of sperm whales in the similar ecosystems ranging from 0.05 yr-1  to 0.09 yr-1 (Evans 1998; Perry et al. 1999). Food web models and data for the Caribbean model, J.L. Melgo et al.  76 In the Caribbean region and surrounding areas, sperm whales were documented to feed primarily on cephalopods species (99.6%); and rarely on crustaceans and benthos (0.4%) and other offshore pelagics (<0.01%) (Table 14). These diet values were weighted based on the estimated consumption of each species of sperm whales. Table 14. Available information on the diet composition of sperm whales used in the Ecopath model for the Caribbean ecosystem. 17 24 25 Source Whale species Areas Other offshore predators Cephalopods Crustace ans and benthos Total dos Santos and Haimovici 2001 Kogia breviceps Southern Brazil  1.000  1.000 Santos et al. 2006 Kogia breviceps Galicia (NW Spain) 0.0003 0.999 0.00008 1.000 Santos et al. 2006 Kogia breviceps France  0.833 0.167 1.000 Santos et al. 2006 Kogia breviceps Scotland (UK)  1.000  1.000 Kawakami 1980 Physeter macrocephalus Peru  1.000  1.000 Kawakami 1980 Physeter macrocephalus Chile  1.000  1.000 Clarke et al. 1980 Physeter macrocephalus Brazil  1.000  1.000 Pascoe et al. 1990 Physeter macrocephalus Patagonia, Argentina  1.000  1.000 Smith and Whitehead 2000 Physeter macrocephalus Galapagos, Ecuador  1.000  1.000 Hickmott 2005  Physeter macrocephalus Northern Bahamas  0.977 0.023 1.000 Weighted average diet for Sperm whales <0.0001 0.996 0.004 1.000  8. Killer whales The species Feresa attenuata (pygmy killer whale), Orcinus orca (killer whales) and Pseudorca crassidens (false killer whale) were combined into single killer whales group. All three species occur throughout the Gulf of Mexico and around the Lesser Antilles (Klinowska 1991; Mohammed et al. 2007b). Debrot et al. (1998) reported that killer whale species occur almost throughout the year in the southeastern Caribbean. The combined annual abundance of all three species in this group in the Caribbean region was estimated to be 1,076 individuals (or 0.06 individuals per 1000 km2) with a total annual biomass of 574 tonnes or an annual biomass density of 0.0003 t*km-2 based on the quantitative estimates by Kaschner (2004). The annual catch density of killer whale species in the region was 0.00001 t*km-2 (Mohammed et al. 2007a). Predicted density (individuals per 1000 km2) was similar in terms of magnitude as observed densities reported from dedicated marine mammals surveys conducted in other areas with similar habitats (Table 7). The P/B ratio input we used for killer whales was 0.02 yr-1, the same value that was used by Trites and Heise (1996), Sidi and Guénette (2004); and Mohammed et al. (2007b) for killer whales in other Ecopath models. Diet information for killer whales diet information was obtained from the Strait of Magellan, Tierra Del Fuego (Alonso et al. 1999), Icelandic waters (Sigurjónsson and Víkingsson 1997) and Lesser Antilles pelagic waters (Heileman et al. 2007). The diet composition of killer whales consisted of 36.5% cephalopods, 26.4% small and schooling pelagics, 15.0% small cetaceans, 14.6% other offshore pelagics and 7.5% zooplankton. These diet proportions were adjusted based on the esimtated consumption of each killer whale species (Table 15). Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  77 Table 15. Diet composition of killer whales used for the Ecopath model in the Caribbean ecosystem. 10 17 21 24 26 Source Whale species Areas Small cetaceans Other offshore predators Small and schooling pelagics Cephalo -pods Zooplan kton Total Alonso et al. 1999 Pseudorca crassidens Tierra Del Fuego (Chile)  0.249  0.751  1.000 Sigurjónsson and Víkingsson 1997 Orcinus orca Icelandic waters   1.000   1.000 Heileman et al. 2007 Orcinus orca, Feresa attenuata Lesser Antilles 0.300 0.15 0.100 0.300 0.150 1.000 Weighted average diet for Killer whales 0.15 0.146 0.264 0.365 0.075 1.000  9. Beaked whales Three species of beaked whales were included in this group: Blainville’s (Mesoplodon densirostris), Gervais’ (Mesoplodon europaeus) and Cuvier’s (Ziphius cavirostris) beaked whales. Beaked whale species are known to occur widely throughout the Caribbean ecosystem in spring and summer months (Debrot et al. 1998). The total annual abundance in the Caribbean ecosystem was estimated to be 264 beaked whales amounting a total annual biomass of 168 tonnes or an average annual biomass density of 0.00013 t*km-2 (Kaschner 2004). The P/B ratio of 0.036 yr-1 was used in the present model, based on what was used by Guénette et al. (2006) in Alaskan ecosystem. Diet information of beaked whale species were obtained from whales strandings in Curacao (Debrot and Barros 1992), western Mediterranean (Blanco and Raga 2000), Northwest Spain and Scotland (Santos et al. 2001), Northern Bahamas (Hickmott 2005) and Canary Island (Santos et al. 2007). All prey items of beaked whales taken from different sources were weighted based on the estimated consumption of each species of beaked whales. These weighted prey items of beaked whales’ species were averaged, and then summed to represent the prey items of our three beaked whale’s species. As a result, we found that beaked whales in the Caribbean fed mostly on cephalopods (68.7%) followed by other offshores pelagics (17.8%), crustaceans and benthos (8.9%) and zooplankton (4.6%) (Table 16). Table 16. Available information on the diet composition of beaked whales: Cuvier’s beaked whales, Blainvilles beaked whales and Gervais beaked whales, in the Caribbean region. A weighted average diet (based on the percentage of food consumption by each species) was used in the Ecopath model for the Caribbean ecosystem. 17 24 25 26 Source Whale species Areas Other offshore predators Cephalopods Crustaceans and benthos Zooplankton Total Hickmott 2005 Mesoplodon densirostris Northern Bahamas 0.500 0.500     1.000 Santos et al. 2007 Mesoplodon densirostris Canary Island  0.848 0.152   1.000 Debrot and Barros 1992 Mesoplodon europaeus Curacao  0.124 0.433  0.443 1.000 Blanco and Raga 2000 Ziphius cavirostris Western Mediterranean  1.000   1.000 Santos et al. 2001 Ziphius cavirostris Nortwest Spain  1.000   1.000 Santos et al. 2001 Ziphius cavirostris North Uist (Scotland)  1.000   1.000 Hickmott 2005 Ziphius cavirostris Northern Bahamas  0.316 0.684  1.000 Santos et al. 2007 Ziphius cavirostris Canary Island   0.9999 0.00014  1.000 Weighted average diet for Beaked whales 0.178 0.687 0.089 0.046 1.000  Food web models and data for the Caribbean model, J.L. Melgo et al.  78 10. Small cetaceans A total of 13 species of small cetaceans were included and aggregated into one group in our model (Table 1). These were Delphinus capensis, Delphinus delphis, Globicephala macrorhynchus, Grampus griseus, Lagenodelphis hosei, Sousa teuszii, Stenella attenuata, S. clymene, S. coeruleoalba, S. frontalis, S. longisrostris, Steno bredanensis and Tursiops truncatus. Some of these species migrate seasonally into the waters of the Caribbean ecosystem, but many can be found in the area throughout the year (Debrot et al. 1998; Mohammed et al. 2007b). Their combined annual average abundance was estimated to be 115,963 individuals amounting to a total annual biomass of 10,467 tonnes and an average annual biomass density of 0.0053 t*km-2 based on the quantitative estimates by Kaschner (2004). The small cetacean species are caught rarely for fisheries in the region with an annual catch density of 0.00002 t*km-2 (Mohammed et al. 2007a) The production to biomass ratio of small cetaceans in the Caribbean ecosystem is 0.03 yr-1 (Mohammed 2003a). This value was incorporated in the present model and was assumed to be representative of small cetaceans P/B ratio for the entire Caribbean ecosystem. Diet information for small cetaceans was derived from the results of Lesser Antilles pelagic ecosystem (Heileman et al. 2007). Diet inputs in the latter study were originally obtained from the marine mammal’s diet composition study by Pauly et al. (1998b) (Table 17). Heileman et al. (2007) weighted these diet compositions by small cetaceans biomass sighted during their Lesser Antilles cetacean survey. In the same study, an import value in small cetacean’s diet was assumed, suggesting that this group feed outside the Lesser Antilles system. However, in our model, we assumed that small cetaceans feed within the Caribbean region and thus, its diet import value from Heileman’s et al. (2007) was redistributed equally to the prey trophic groups of small cetaceans. As a result, small cetaceans diets were composed of 58.73% cephalopods, 37.04% other offshore predators, 4.22% small and schooling pelagics and 0.01% small cetaceans. Table 17. Available information on the diet composition of small cetaceans in the Caribbean region. A weighted average diet (based on the percentage of food consumption by each species) was used in the Ecopath model for the Caribbean ecosystem. 17 22 25 26 Source Areas Small cetaceans Other offshore pelagics Small and schooling pelagics Cephalopods Total Heileman et al. 2007 Lesser Antilles 0.0001 0.3704 0.0422 0.5873 1.000  Food consumption by marine mammals Kaschner (2004) developed a basic food consumption model based on Trites et al. (1997). This model was used to generate the biomass and consumption (Q/B) values needed for each Ecopath group. Annual food consumption was calculated as: != s s,is,ii RN*365Q where Q of species i was assumed to be 365 times the daily food consumption. Daily food consumption was calculated based on the number of individual N of sex s of a species i, and a weight-specific daily ration R calculated based on the mean individual body mass W of sex s belonging to species i, consumed by each individual of species i and sex s. Abundances and sex ratios were taken directly from the Kaschner (2004) database. Mean species and sex-specific body mass was taken from Trites and Pauly (1998). For all cetaceans, except baleen whales, we used the empirical model developed by Innes et al. (1987) to estimate food consumption of cetaceans that was later modified by Trites et al. (1997) to account for the difference between consumption for growth and for maintenance and then applied to all marine mammal species. Food intake of specific species per day was calculated using: Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  79 8.0 s.isi, W*0.1R = where R is the daily food intake of an individual of sex s belonging to species i and W  is the mean body weight of that individual, in kilograms. For all baleen whales daily food ration was estimated based on a model by Armstrong and Siegfried (1991) for food consumption of minke whales in the Antarctic. These authors suggested a modification to the empirical model of Innes et al. (1986) equation for baleen whales to account for larger body sizes and seasonal variation in food intake. This approach was later used to estimate food consumption of whales around Iceland (Sigurjónsson and Víkingsson 1997) and represents one of the methods used by Tamura (2003) to estimate global food intake of cetaceans. This feeding rate is calculated as: ! R i,s = 0.42*W i,s 0.67 Annual food consumption for each species of marine mammals was then divided by the biomass estimates, in order to get the final consumption to biomass (Q/B) ratios used in the Ecopath model Table 18).  Table 18. Consumption estimates for each marine mammal groups of the Ecopath model of the Caribbean. Ecopath groups Annual food consumption (tonnes*km-2) Annual biomass (tonnes*km-2) Q/B Minke whales 0.0381 0.0045 8.421 Fin whales 0.0257 0.0062 4.161 Humpback whales 0.8470 0.1667 5.081 Bryde’s whales 0.0270 0.0043 6.260 Sei whales 0.0053 0.0009 6.178 Blue whales 0.0084 0.0025 3.398 Sperm whales 0.0947 0.0188 5.030 Killer whales 0.0028 0.0003 9.468 Beaked whales 0.0013 0.0001 9.933 Small cetaceans 0.0769 0.0053 14.404  11. Seabirds Twenty one species of seabirds were included in this model and pooled into one group (Table 1). Most of these seabirds’ species are commonly found nesting in the Bahama archipelago, Greater and Lesser Antilles and Trinidad and Tobago (Mohammed et al. 2007b). Seabirds population in the West Indies occur in relatively low density and are often found in offshore rock or inaccessible cliffs (Schreiber and Lee 2000; Mohammed et al. 2007b). Most of their habitats or nesting areas have been converted into coastal developments (Mohammed et al. 2007b). Only few information of the population status of seabirds in the Caribbean exist (Mohammed et al. 2007b). Hence, the seabird species considered in our model are restricted to species that has trophic impact on the pelagic fish species, following the Lesser Antilles pelagic ecosystem model (Mohammed et al. 2007b). The annual biomass density of 0.0002 t*km-2 seabirds in Lesser Antilles pelagic ecosystem model by Mohammed  et al. (2007b) was based on the product of the number of nesting pairs in the West Indies (Schreiber and Lee 2000) and individual weights of seabirds (Vasconcellos and Watson 2004; Mackinson et al. 2005; Priddel et al. 2005). The latter value was then incorporated in the present model and assumed to be representative for the Caribbean ecosystem. The seabirds P/B ratio of 0.13 yr-1 used in the present model for Caribbean ecosystem was obtained from the Lesser Antilles pelagic ecosystem model (Mohammed et al. 2007b). This was assumed to be Food web models and data for the Caribbean model, J.L. Melgo et al.  80 similar P/B ratio for all seabird species in the entire Caribbean ecosystem.  Likewise, Q/B estimate of the similar species for seabirds used in the Caribbean ecosystem model was taken from Lesser Antilles pelagic ecosystem model (Mohammed et al. 2007b). This Q/B value was 73.690 yr-1 and assumed to be representative in the entire Caribbean ecosytem. The diet information for seabirds in the present model was obtained from Opitz (1996). The diet of seabirds is composed of other offshore predators, reef fishes, coastal predators, crustaceans and benthos and zooplankton (Opitz 1996).  12. Seaturtles In the present model, the seaturtles group included were green seaturtle (Chelonia mydas), hawksbill seaturtle (Eretmochelys imbricate), leatherback turtle (Dermochelys coriacea), and loggerhead seaturtle (Caretta caretta). The leatherback and hawksbill turtles are critically endangered species, and the green turtle is considered to be a threatened species (Grazette et al. 2007). These species are nesting in the Caribbean, from Costa Rica to Colombia, from French Guiana to Surinam, along the central Brazilian coast, Guyana, Trinidad, the Dominican Republic, Virgin Islands, Puerto Rico and along the western coast of Mexico to Panama (Grazette et al. 2007; Read et al. 2007).  The seaturtles population in the Caribbean ecosystem are threatened from coastal and upland developments, introduction of domestic and nonindigenous animals, boating (both commercial and recreational), incidental capture in fisheries and illegal harvest of adults and eggs (Bell et al. 2007; Mohammed et al. 2007b). Their aggregated annual biomass in the Lesser Antilles pelagic ecosystem was estimated to 0.001 t*km-2 (Mohammed et al. 2007b). Other published Ecopath models used annual biomass of the similar species for seaturtles ranging from 0.0026 t*km-2 in Eastern Tropical Pacific ecosystem (Olson and Watters 2003) to 0.070 t*km-2 in Southern Mexican Caribbean (Alvarez-Hernández 2003). Here, we used mid- range annual biomass value of 0.037 t*km-2 (based on eight ecosystem models) for the present model and assumed to be representative for the Caribbean ecosystem. Seaturtles are traditionally harvested in the Caribbean ecosystem (Bell et al. 2007; Grazette et al. 2007, Mohammed et al. 2007b). These animals are a valuable fishery resource for meat (Finlay 1984), shells sold in local and formerly on international market (Eckert and Eckert 1990). Their eggs are also considered a traditional delicacy in some areas in the Caribbean (Grazette et al. 2007). Seaturtles are caught with harpoons, spearguns and sometimes nets (Grazette et al. 2007; Mohammed et al. 2007b). The annual average catch density of seaturtles for local fleets in the Caribbean was 0.00003 t*km-2 (Mohammed et al. 2007b). There was no information about foreign fleets’ catches of this group in the area. Of all seaturtle species included herein, it is the leatherback turtle that are commonly caught in region (Mohammed et al. 2007a). The P/B ratio of 0.15 yr-1 for seaturtles, used in the Lesser Antilles pelagic ecosystem was originally derived from Opitz (1996) for the Caribbean coral reefs ecosystem. In the other published models for the Caribbean areas, this parameter ranged from 0.15 yr-1 in Eastern Tropical Pacific (Olson and Watters 2003) to to 1.52 yr-1 in Southern Mexican ecosystem Caribbean (Alvarez-Hernández 2003). Based on the eight published ecosystem model, we used mid-range value of 0.835 yr-1 in the present model representing the seaturtle P/B ratio for the entire Caribbean ecosystem. The Q/B value of the similar species for seaturtles used in the Lesser Antilles pelagic ecosystem was 3.500 yr-1 (Mohammed et al. 2007b).  This value was originally obtained from the Caribbean coral reef ecosystem model by Opitz (1996). Additionally, the latter value was the minimum range of the Q/B values in the other study areas of the Caribbean that ranged from 3.500 yr-1 in Eastern Tropical Pacific (Olson and Watters 2003) to 3.570 yr-1 in the Caribbean coast of Southern Mexican ecosystem (Alvarez- Hernández 2003). In the present model, we used mid-range Q/B value of 3.535 yr-1 (based on eight published ecosystem models) representing the Caribbean ecosystem. Seaturtle diet information inputs in the present model were obtained from Opitz (1996). According to Opitz  (1996), seaturtle prey on small and schooling pelagics, reef fishes and benthic producers.  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  81 13. Large tunas and billfishes In the initial model, of albacore (Thunnus alalunga), bigeye tuna (T. obesus), yellowfin tuna (T. albacares), swordfish (Xiphias gladius) and billfishes (Istiophorus sp. Tetrapturus albidus,) were treated as separate trophic groups (Mohammed et al. 2007b). However, in the present model, these trophic groups were aggregated into one group as large tunas and billfishes. Likewise, their Ecopath parameter values were aggregated as well. The large tuna and billfish species are commercially important fish species migrating along Caribbean, Western Atlantic and Eastern Pacific (Mohammed et al. 2007b). The keyspecies of this group were albacore tuna, bigeye tuna, yellowfin tuna, swordfish and billfishes. Their annual biomass around the Lesser Antilles pelagic ecosystem ranged from 0.0006 t*km2 to 0.0120 t*km2 and was 0.0272 t*km2 after aggregating process (Mohammed et al. 2007b). The latter value was incorporated in the present model and assumed to be representatitive in the Caribbean ecosystem. The large tunas and billfishes are overfished in the area (Mohammed et al. 2007a). The annual average catch density for local fleets was 0.00467 t*km-2 and was 0.0002 t*km-2 for foreign fleets (Mohammed et al. 2007a; Sea Around Us database unpublished data from Reg Watson and www.seaaroundus.org). This group is usually exploited by locals during fishing trips using boats and canoes with outboard engines (Mahon 1990; Mohammed et al. 2007a; Mohammed et al. 2007b). Some have larger inboard powered launches with ice containers (Mahon 1990). The yellowfin tuna is the most important and targeted large pelagic species in the Caribbean Sea and in the Atlantic Ocean (Marcano et al. 2004; Mohammed et al. 2007a; Mohammed et al. 2007b). The key landing sites for billfishes are in Barbados, Grenada, St. Lucia and St. Vincent, and the Grenadines (Mahon et al. 1994a). In addition, Restrepo et al. (2003) and Marcano et al. (2004) suggested that large tuna’s species are highly targeted by recreational or artisanal fisheries, and some are bycatch of tuna longline operations. The P/B ratios of large tunas and billfishes group in the initial model were derived from the information on the natural and fishing mortalities of the similar species used in ICCAT (2004) stock assestments and from Pauly (1980). This P/B estimates ranged from 0.37 yr-1 to 2.0 yr-1 in the Lesser Antilles pelagic ecosystem, and was 1.25 yr-1 after aggregation (Mohammed et al. 2007b).  The latter value was then used in the present model representing the P/B ratio for large tunas and billfishes in the entire Caribbean ecosystem. The aggregated Q/B value of 15.530 yr-1 for the similar species of large tunas and billfishes from the Lesser Antilles pelagic ecosystem model was used in the present model representing the Caribbean ecosystem. This Q/B value in Lesser Antilles pelagic ecosystem could range from 4.596 yr-1 to 17.590 yr-1 (Mohammed et al. 2007b). Large tunas and billfishes group preys mostly on large tunas (juveniles), small tunas, dolphinfish, other offshore predators, flyingfish, scombrids, small and schooling pelagics, coastal predators, cephalopods and zooplankton (Heileman et al. 2007).  14. Small tunas The small tunas group was mainly composed of Atlantic skipjack tuna (Katsuwonus pelamis), blackfin tuna (Thunnus atlanticus), bullet tunas (Auxis sp.), Atlantic bonito (Sarda sarda) and little tunny (Euthynnus alletteratus). Originally, in the initial model, this group was separated into two trophic groups; the skipjack tuna and other offshore predators group. In the present model, these two trophic groups were aggregated into one group as small tunas, in order to reduce the complexity of the model. In general, there is limited information on the stock structure and abundance of small tuna species except for Atlantic skipjack tuna (Mohammed et al. 2007b). This species is found in the Caribbean, off Brazil, in the Gulf of Mexico, and in the Gulf of Guinea for their spawning season during summer months (Fonteneau and Marcille 1993). The skipjack tuna occurring in the Caribbean region is considered part of the western Atlantic stock (Mohammed et al. 2007b). Since skipjack tuna is an important species for the small tuna trophic group in the area, we decided to use the Ecopath parameter values of this species obtained from Mohammed et al. (2007b) as inputs for the present model. These inputs were assumed to be representative in the Caribbean ecosystem. Food web models and data for the Caribbean model, J.L. Melgo et al.  82 The annual biomass, P/B and Q/B estimates used in the present model for the Caribbean ecosystem were 0.0119 t*km-2, 1.96 yr-1 and 19.610 yr-1, respectively. The annual average catch density of small tunas for local and foreign fleets in the Caribbean was 0.00678 t*km-2 and 0.00005 t*km-2, respectively (Mohammed et al. 2007a; Sea Around Us database unpublished data from Reg Watson and www.seaaroundus.org). This group is targeted for both recreational and artisanal fishing in some areas in the Caribbean (Restrepo et al. 2003). According to Heileman et al. (2007), small tunas mainly feed on small tuna species, small and schooling pelagics and cephalopods. This species rarely feed on dolphinfish, flyingfish, coastal predators, scombrids and zooplankton.  15. Dolphinfish The common dolphinfish, Corypahena hippurus, and the pompano dolphinfish, C. equiselis compose the dolphinfish group in our model. This group was retained herein as a separate group because of its significant importance to both commercial and recreational fisheries for large pelagic fishes in the Western Central Atlantic as well as in the Caribbean ecosystem (Mahon 1996; Mahon 1999; Oxenford 1999; Parker et al. 2001; Mohammed et al. 2007b). Dolphinfish species are migratory fish species that occur mostly in Southeastern US, Bermuda, Lesser Antilles areas and in north coast of Brazil (Oxenford and Hunte 1986; Mahon 1999; Mahon and Oxenford 1999; Die 2004). According to the genetic study of this group by Die (2004), the dolphinfish stock in the Caribbean belongs to a single stock in the Western Central Atlantic. The annual biomass of this group was obtained from the Lesser Antilles pelagic ecosystem model developed by Mohammed et al. (2007b). Dolphinfish annual biomass was 0.0278 t*km-2 and assumed to be representative for the Caribbean ecosystem. The status of dolphinfish stock in the Caribbean is highly uncertain (Mohammed et al. 2007a). Some experts (e.g. Prager 2000; Die 2004) suggest that the stock is not overfished, while others (e.g.Parker et al. 2001) show that the dolphinfish stock is under intense overfishing. According to Mahon (1999), distant water fleets fishing in the Caribbean areas do no report any dolphinfish landings, in particular recreational fisheries and bycatch in large-scale commercial fisheries, resulting to an unclear decline status of this resource in the area. Based on the the timeseries catch densities results of this group from Mohammed et al. (2007a) and Sea Around Us database (unpublished data from Reg Watson and www.seaaroundus.org), it showed that the annual average catch density of dolphinfish in the region was 0.00512 t*km-2 and 0.00003 t*km-2 for local and foreign fleets, respectively. Similar to the large tunas and billfishes group, the dolphinfish stocks are widely exploited in the region as well (Mahon 1990; Mahon and Oxenford 1999). They are caught by a wide range of fisheries: artisanal, small-scale commercial, large scale-commercial and recreational (Mahon 1999; Mohammed et al. 2007a). The P/B ratio of 4.72 yr-1 of dolpinfish used in the present model for Caribbean ecosystem was obtained from Mohammed et al. 2007b) for the Lesser Antilles pelagic ecosystem model. This value was the mid- range value used in the Lesser Antilles based from the reported total mortality rates (equivalent to P/B ratio) in the same study area, ranged from 3.53 yr-1  to 8.67 yr-1 (see Oxenford 1985; Murray 1985; Bentivoglio 1988; Parker et al. 2001). The dolphinfish Q/B input in Lesser Antilles was 20.000 yr-1 (Mohammed et al. 2007b). The latter value was directly used in the present model and assumed to be representative in the entire Caribbean ecosystem. Generally dolphinfish are piscivorous, eating a wide variety of fish species including large tunas and billfishes, small tunas, other offshore predators, flyingfish, scombrids and coastal predators (e.g. jacks) (Oxenford and Hunte 1999; Heileman et al. 2007). They also prey on cephalopods, crustaceans and benthos and zooplankton (Heileman et al. 2007).  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  83 16. Flyingfish Flyingfish species are an important component of pelagic fisheries in the Southeastern Caribbean (Mahon 1990; Mohammed et al. 2007a). Specifically, the fourwing flyingfish, Hirundichthys affinis, contributes 95% of the fishery catches in the Lesser Antilles (Mahon 1989; Mohammed et al. 2007b). Other most common flyingfish species in the Southeastern Caribbean are margined flyingfish (Cheilopogon cyanopterus) and sailfin flyingfish (Parexocoetus brachypterus) (Mahon 1989). Because of the high importance of these three flying fishes, they were considered as the representative for the flyingfish group for the Caribbean ecosystem model. The biomass of flyingfish in Mohammed’s et al. (2007b) model for the Lesser Antilles pelagic ecosystem was derived from population surveys of the species in the area conducted by Oxenford et al. (1995). As a result, the annual biomass of flyingfish group in the area was 0.2080 t*km-2 (Mohammed et al. 2007b). This value was also incorporated in present model and assumed to be representative in the Caribbean ecosystem. The annual average catch density of this group for local fleets in the Caribbean was 0.00288 t*km-2 (Mohammed et al. 2007a). There was no catches of flyingfish in the foreign fleets. The most targeted flyingfish species was fourwing flyingfish, H. affinis that contributed almost 95% of the landings in the Lesser Antilles (Mohammed et al. 2007b). Additionally, flyingfish were mainly caught by trolling boat with dipnets in St. Vincent and the Grenadines and by outboard fleets using gillnets in Grenada (Mohammed et al. 2007a). The P/B ratio of 4.0 yr-1 for flyingfish group used in the present model for the Caribbean ecosystem was obtained from the Lesser Antilles pelagic ecosystem model developed by Mohammed et al. (2007b). Additionally, the used P/B value from Lesser Antilles was derived total mortalities values and overall catches percentages of the keyspecies of the group: fourwing flyingfish, margined flyingfish and sailfin flyingfish (Mohammed et al. 2007b).  Likewise, the Q/B estimate of this group was obtained Lesser Antilles pelagic ecosystem model. The value was 24.760 yr-1 (Mohammed et al. 2007b). The flyingfish species found in the region feed mainly on of small and schooling pelagics and zooplankton (Heileman et al. 2007).  17. Other offshore predators The “other offshore predators”: group was described as fish species that was found in mesopelagic all the way to the deep layers of the ocean. Originally, the offshore predator in the initial model was composed of small tuna species and oceanic triggerfish. However, since we established a separate trophic group for small tunas, we decided to aggregate oceanic triggerfish and small to large mesopelagic fish from LAPE model into a group for “other offshore predators”. Other fish species belonging to this group were Alepocephalidae, Gonostomatidae, Macrouridae, Moridae and Myctophidae (Table 1). The annual biomass of the similar species of this group in Lesser Antilles ranged from 8.7240 t*km-2 to 10.8330 t*km-2 (Mohammed et al. 2007b). This estimate was derived from hydroacoustic and pelagic trawl surveys in Lesser Antilles pelagic ecosystem by Melvin et al. (2007). The ranges of annual biomass of other offshore predators in other ecosystem models in adjacent Caribbean ranged from 0.0006 t*km-2 to 3.2540 t*km-2, which were both used in Central Atlantic ecosystem (Vasconcellos and Watson 2004). In the present model, we used an annual biomass of the similar species for other offshore predators of 1.627 t*km-2 and assumed to be representative in the Caribbean ecosystem. The latter value was a mid-range estimate of the similar species used in the other published ecosystem models in the adjacent Caribbean areas. The annual catch density of the similar species for other offshore predators group for local fleets was as little as 0.00000005 t*km-2 (SAUP unpublished data from Reg Watson and www.seaaroundus.org). The annual catch of this group for foreign fleets was 0.00008 t*km-2 (Mohammed et al. 2007a; SAUP unpublished data from Reg Watson and www.seaaroundus.org). This group is caught on the same fishing trips with dolphinfish, flyingish, and large tunas and billfishes (Mahon 1990). Food web models and data for the Caribbean model, J.L. Melgo et al.  84 The P/B ratio, ranged from 0.355 yr-1 to 3.76 yr-1, for other offshore group in Lesser Antilles was based on the mean habitat temperature (150C) and growth parameters from FishBase (www.fishbase.org; Mohammed et al. 2007b). This P/B estimates were then calculated using the empirical equation after Pauly (1980). In other Ecopath models in adjacent Caribbean ecosystem, the P/B ratio of the similar species for other offshore predators ranged from 0.15 yr-1 to 3.757 yr-1 in Central Atlantic (Vasconcellos and Watson 2004). We used herein mid-range P/B estimate of 1.863 yr-1 (based on eight ecosystem models) and assumed to be representative for the Caribbean ecosystem. The Q/B value of other offshore predators in Lesser Antilles was 3.550 yr-1 to 15.000 yr-1 (Mohammed et al. 2007b). This was calculated based on the mean habitat temperature and on asymptotic weight and aspect ratios from Fishbase (www.fishbase.org) using the the empirical equation after Pauly (1980) (Mohammed et al. 2007b).  The other Q/B estimates of the similar species for other offshore predators in adjacent Caribbean ranged from 0.290 yr-1 to 18.250 yr-1 in Central Atlantic (Vasconcellos and Watson 2004). In the present model, a mid-range value of 9.270 yr-1 was used representing the entire Caribbean ecosystem. According to Heileman et al. (2007), the other offshore predators feed mainly on benthic producers.  18. Pelagic sharks Various sharks’ species were included in this group (Table 1). The most important species of the group were the blue shark (Prionace glauca), longfin mako (Isurus paucus), oceanic whitetip (Carcharhinus longimanus), porbeagle (Lamna nasus), shorfin mako (I. oxyrinchus), spinner shark (C. brevipinna) and tiger shark (Galeocerdo cuvier).  The pelagic sharks are distributed throughout the Atlantic Ocean (Mahon 1990). The annual biomass of this group used in the present model for the Caribbean ecosystem was 0.0116 t*km-2. The latter value was derived from the annual biomass of the same group in the Lesser Antilles pelagic ecosystem model by Mohammed et al. (2007b). The estimate further represents an intermediate value of the possible annual biomass ranges of the similar species for pelagic sharks from other ecosystem models in adjacent areas of the Caribbean, ranging from 0.0004 t*km-2 in Eastern Tropical Pacific (Olson and Watters 2003) to 0.0300 t*km-2 in Bahia Ascension, Mexican Caribbean (Vidal and Basurto 2003). The timeseries catch density for pelagic sharks in the Caribbean ecosystem were taken from the Lesser Antilles national fisheries report (Mohammed et al. 2007a) and from Sea Around Us Database (unpublished data from Reg Watson and www.seaaroundus.org). The average annual catch density for pelagic sharks was 0.00025 t*km-2 for local fleets and as low as 0.00005 t*km-2 for foreign fleets (Mohammed et al. 2007a; Sea Around Us Database unpublished data from Reg Watson and www.seaaroundus.org). Pelagic shark species species were taken as bycatch and some are targeted in commercial fisheries using bottom gillnet and long line at certain times of the year (Chan and Shing 1999; Mohammed et al. 2007a). The P/B ratios used for pelagic sharks in the Caribbean ecosystem model was 0.4 yr-1. This value was obtained from the same group in the Lesser Antilles pelagic ecosystem model (Mohammed et al. 2007b). Likewise, Q/B value of 10.000 yr-1 for the same group in the Lesser Antilles pelagic ecosystem model (Mohammed et al. 2007b) was incorporated in the present model. This was assumed to be representative for the entire Caribbean ecosystem. Pelagic sharks largely feed on small cetaceans, seaturtles, large tunas and billfishes, small tunas, dolphinfish, other offshore pelagics, pelagic sharks, scombrids, small and schooling pelagics, reef fishes, coastal predators, cephalopods, zooplankton and detritus (Opitz 1996; Heileman et al. 2007).  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  85 19. Coastal and demersal sharks and rays The coastal sharks and rays included herein are shown in Table 1. The important species of the group were nurse sharks (Ginglymostoma cirratum), rays (Myliobatidae) and southern stingray (Dasyatis americana). The coastal sharks and rays group was not included in the initial model since their interest focussed mainly on the pelagic ecosystem in Lesser Antilles. In the present model, we cover all areas in the Caribbean including the coastal habitats of marine organisms in the region. Hence, we decided to add the group of sharks and rays inhabiting the coastal zone in the Caribbean ecosystem. Coastal sharks are occasionally present in the region, while rays are fairly common in the coral reefs (Opitz 1996). The annual biomass of the similar species for coastal and demersal sharks and rays in the adjacent Caribbean ecosystem ranged from 0.0002 t*km-2 in Eastern Tropical Pacific (Olson and Watters 2003) to 0.4000 t*km-2 in Southern Mexican Caribbean (Alvarez-Hernández 2003). Here, a mid-range annual biomass estimate of 0.2000 t*km-2 was used in the present model and assumed to be representative in the entire Caribbean ecosystem. The coastal and demersal sharks and rays are harvested in region (Mahon 1990; Opitz 1996). Some of the species of this group are incidentally caught in trap or handline fisheries on the island shelves (Mahon 1990). The most common shark species caught in the latter gear are nurse sharks (G. cirratum) (Mahon 1990). Based from the LAPE national fisheries (Mohammed et al. 2007a) and Sea Around Us database (unpublished data from Reg Watson and www.seaaroundus.org), the annual average catch density of this group for local fleets and foreign fleets were 0.000177 t*km-2 and 0.00018 t*km-2, respectively. The P/B estimates of the similar species for this group in the adjacent Caribbean could range from 0.112 yr-1 in Central Atlantic (Vasconcellos and Watson 2004) to 0.6 yr-1 in Southeastern Caribbean (Mohammed 2003a). In the present model, a P/B estimate of 0.356 yr-1 was used representing the Caribbean ecosystem. This value was mid-range P/B estimate of the similar species used in eight ecosystem models in the adjacent Caribbean ecosystem. The coastal sharks and rays Q/B estimates in the Caribbean areas ranged from 1.800 yr-1 in Central Atlantic (Vasconcellos and Watson 2004) to 9.160 yr-1 in Eastern Tropical Pacific (Olson and Watters 2003). Here, a mid-range value of 5.480 yr-1 was used in the present model and assumed to be representative in the Caribbean ecosystem. Sharks and rays largely feed on variety of preys such as seaturtles, small tunas, dolphinfish, coastal and demersal sharks and rays, small and schooling pelagics, scombrids, coastal predators, reef fishes, cephalopods, crustaceans and benthos, and benthic producers (Opitz 1996; Heileman et al. 2007).  20. Scombrids In the initial model, scombrids species such as wahoo and other mackerels were grouped separately. However, we decided to aggregate all the scombrids species in order to reduce the complexity of the present model for the Caribbean ecosystem. The important species of the scombrids group were cero (Scomberomorus regalis), king mackerel (S. cavalla), Spanish mackerel (S. brasiliensis) and wahoo (Acanthocybium solandri). These species are migratory and occupy US coastal waters of the South Atlantic and the Gulf of Mexico (Mahon and McConney 2004). Their annual biomass in the Lesser Antilles areas ranged from 0.0010 t*km-2 to 0.0650 t*km-2, and was 0.0660 t*km-2 after aggregation (Mohammed et al. 2007b). These biomass estimates were based on the ratio of catches inside the Lesser Antilles pelagic ecosystem and the total catch (Mohammed et al. 2007b). The aggregated annual biomass (0.0660 t*km-2) of this group from LAPE model was used directly in the present model for the Caribbean ecosystem. The scombrids species are commercially important component in the Caribbean fisheries (Mahon 1990; Optiz 1996; Oxenford et al. 2003; Mohammed et al. 2007ab). These fish species are also targeted for recreational fishing in the area (Mohammed et al. 2007b). The average annual catch density of scombrids in the area was 0.01129 t*km-2 for local fleets and 0.00079 t*km-2 for foreign fleets (Sea Food web models and data for the Caribbean model, J.L. Melgo et al.  86 Around Us database unpublished data from Reg Watson and www.seaaroundus.org). The most harvested scombrids in the Caribbean ecosystem are king mackerel, Spanish mackerel and wahoo (Mahon 1990; Mohammed et al. 2007a). The P/B ratio used in the present model for the Caribbean ecosystem was 1.09 yr-1. This value was obtained from the P/B ratio of mackerel/scombrids group in the Lesser Antilles pelagic ecosystem (Mohammed et al. 2007b). The Q/B value of 10.310 yr-1 incorporated in the present model was derived from the Q/B estimate of the mackerel group in the Lesser Antilles pelagic ecosystem (Mohammed et al. 2007b). This value seemed to be the mid-range of the possible Q/B estimates of the similar species for scombrids in the other ecosystem models in adjacent areas in the Caribbean, ranging from 9.150 yr-1 in Caribbean coral reefs ecosystem model (Opitz 1996) to 11.400 yr-1 in Eastern Tropical Pacific model (Olson and Watters 2003). Diet information of scombrids was obtained from Heileman et al. (2007). From their study, it showed that scombrids preyed mainly on flyingfish, coastal predators, cephalopods and zooplankton, and occasionally feed on small tunas, scombrids, and small and schooling pelagics.  21. Small and schooling pelagics In this model, the small coastal pelagics and small offshore pelagic fish trophic groups from Mohammed et al. (2007b) version were pooled into ‘small and schooling pelagics’ because of their similarities in ecological importance in the study area, and also to have a simplified model structure for the Caribbean ecosystem model. This group represents various species of small and schooling pelagic included in the model: anchovies, halfbeaks, herrings, scads and small jacks (Table 1). The key species of this group were anchovies (Anchoa sp.), Atlantic bumper (Chloroscombrus chrysurus), bigeye scad (Selar crumenophthalmus), herring (Harengula sp.), sardine (Sardinella sp.), scad (Decapterus sp.), and threadfin shad (Dorosoma petenense). The aggregated annual biomass of small and schooling pelagics in Mohammed et al. (2007b) model for the Lesser Antilles pelagic ecosystem was 0.2440 t*km-2 to 7.3940 t*km-2 and was 10.1970 t*km-2 after aggregating process (Mohammed et al. 2007b). Other ecosystem models in adjacent areas of the Caribbean used annual biomass values that ranged from 0.1500 t*km-2 in Bahia Ascencion, Mexican Caribbean (Vidal and Basurto 2003) to 33.0000 t*km-2 in Southeastern Caribbean (Mohammed 2003a). In the present model, we used an intermediate or mid-range annual biomass value of 16.575 t*km-2 for the similar species of small and schooling pelagic in the Caribbean ecosystem. The average annual catch density for local and foreign fleets of small and schooling pelagics was 0.03478 t*km-2 and 0.0022 t*km-2, respectively (Mohammed et al. 2007a; Sea Around Us database unpublished data from Reg Watson and www.seaaroundus.org). The small and schooling pelagics are usually caught by longline fleets in the region (Mohammed et al. 2007a). In addition, catches of most of these species are used for live bait fishing for commercial and recreational fisheries in the region (Heileman-Manickchand 1992). The P/B ratio for small and schooling pelagics groups in the Lesser Antilles pelagic ecosystem model ranged from 3.5 yr-1 to 3.6 yr-1, and was 3.245 yr-1 after the group aggregations (Mohammed et al. 2007b). Other models for adjacent areas of the Caribbean P/B used values from 0.38 yr-1 in the Caribbean coral reef ecosystem model (Opitz 1996) to 5.5 yr-1 in Costa Rica (Wolff et al. 1998). Thus, we used herein an intermediate or mid-range P/B ratio of 2.94 yr-1 and assumed to be representative for the Caribbean ecosystem. The small and schooling pelagics Q/B values in the Lesser Antilles pelagics ecosystem was 14.640 yr-1 after aggregation process (Mohammed et al. 2007b). The Q/B estimates for similar species in other Caribbean ecosystems models could vary from 3.900 yr-1 to 43.400 yr-1 in Caribbean coral reefs ecosystem (Opitz 1996). Here, a mid-range Q/B estimate of 23.650 yr-1 was used for small and schooling pelagics in our model. Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  87 Diet information for small and schooling pelagics was obtained from Heileman et al. (2007). Therein, it was documented that small and schooling pelagics feed mostly on small and schooling pelagics, reef fishes, crustaceans and benthos, and zooplankton.  22. Reef fishes The reef fish trophic group in the initial model was pooled with the coastal predators group. However, because of the importance of reef fishes in some areas (e.g. Virgin Islands, Puerto Rico) in the Caribbean (see Opitz 1996), we decided to make “reef fishes” separate in the present model. The reef fish resources are a major source of fish and fishery products to the domestic food market in many Caribbean islands (Jeffrey 2000).  Many reef fish species are included herein (Table 1). The key species of this group were species belonging to Bermuda chab (Kyphosus sectatrix), grouper (Epinephelus sp.), Haemulidae, queen parrotfish (Scarus vetula) and rainbow parrot fish (Scarus guacamaia). These species inhabits not only the coral reef areas but are also found along the seagrass beds, and its juveniles are often in the mangrove areas (Opitz 1996). According to Mahon (1990), the reef fishes in the region are considered to be overexploited with possible excemption of a few islands with large shelves. The annual biomass of the similar species for reef fish group in other Ecopath models in the adjacent areas in the Caribbean ranged from 0.007 t*km-2 in Southeastern Caribbean (Mohammed 2003a) to 46.688 t*km-2 in Caribbean coral reef ecosystem (Opitz 1996). The extremely high range of annual biomass (99.000 t*km-2) of the similar species in the Southern Mexican Caribbean was not used to factor the mid-range annual biomass value for the reef fishes group. Consequently, we used annual biomass estimate of 0.654 t*km-2 for reef fishes group, representing the Caribbean ecosystem. The latter value was a mid-range estimate of the available annual biomass values of the similar species derived from the eight ecosystem models in the adjacent Caribbean areas. The timeseries catch density of reef fish group from 1987 to 2005 were taken from Lesser Antilles fisheries report (Mohammed et al. 2007a) and in Sea Around Us database (unpublished data from Reg Watson and www.seaaroundus.org). Its annual average catch for both local and foreign fleets in the Caribbean was 0.00792 t*km-2 and 0.00091 t*km-2, respectively. Reef fish species in the Caribbean region are highly targeted for artisanal or inshore fishery operations (Nemeth 2005) and some are used for baits or harvested for aquarium trade (Opitz 1996). Reef fish species belonging to Haemulidae, Holocentridae, Monocanthidae, Mullidae, Muraenidae, and Serranidae are frequently caught in the area (Gobert 2000). These species are caught mostly using traps and handlines (Munro 1983; Recksiek et al. 1991; Mahon 1990; Gobert 2000). Most of the reef fishes in some areas in the Caribbean (e.g. Martinique, Guadeloupe, Puerto Rico) are overfished and heavily exploited (Opitz 1996; Gobert 2000; Jeffrey 2000) The P/B ratios for reef fish groups in the other Ecopath ecosystem models in adjacent areas of the Caribbean ranged from 0.37 yr-1 to 3.14 yr-1. These values minimum and maximum P/B ranges were both originate from the Caribbean coral reefs ecosystem model (Opitz 1996). In the present model, mid- range P/B estimate of 1.755 yr-1 was used, representing the Caribbean ecosystem. In the other study areas of the Caribbean, the Q/B used values of the similar species for reef fishes ranged from 2.300 yr-1 to 39.700 yr-1. These ranges were both originate from the Caribbean coral reef ecosystem (Opitz 1996). Here, a mid-range Q/B estimate of 21.000 yr-1 for reef fishes group, based on eight Caribbean models, was used in the present model representing the Caribbean ecosystem. The diet of the reef fishes group was mainly composed of reef fishes, small and schooling pelagics, cephalopods, crustaceans and benthos, zooplankton, benthic producers and phytoplankton (Opitz 1996).  Food web models and data for the Caribbean model, J.L. Melgo et al.  88 23. Coastal predators The key species of coastal predators group were amberfish (Seriola sp.), common snook (Centropomus undecimalis), jacks (Caranx sp.), leatherjacks (Oligoplites sp.), needle fish (Belonidae), pompanos (Alectis ciliaris, Trachinotus sp.,), rainbow runner (Elegatis bipinnulata) and snapper (Lutjanus sp.). Their annual biomass used in the present model for the Caribbean ecosystem was 1.260 t*km-2. This value was directly derived from the same group in the Lesser Antilles pelagic ecosystem model (Mohammed et al. 2007b). The average annual catch density for local and foreign fleets of coastal predators in the region was 0.00735 t*km-2 and 0.00128 t*km-2, respectively (Mohammed et al. 2007a; Sea Around Us database unpublished data from Reg Watson and www.seaaroundus.org). The snappers and jacks are widely harvested coastal predators species in the Caribbean region (Mohammed et al. 2007a). The P/B ratio input for coastal predators in present model for the Caribbean ecosystem model was 0.72 yr-1. This value was obtained from the same group in the Lesser Antilles pelagic ecosystem model (Mohammed et al. 2007b). Likewise, Q/B input value of 7.220 yr-1 in the present model was obtained from the Lesser Antilles pelagic ecosystem model (Mohammed et al. 2007b).  The latter value was assumed to be representative for coastal predators in the entire Caribbean ecosystem. The diet contents of coastal predators are composed of coastal predator species, reef fishes, small and schooling pelagics, cephalopods, crustaceans and benthos, zooplankton and benthic producers (Opitz 1996).  24. Cephalopods The large and small squid groups in the Lesser Antilles pelagic ecosystem model were pooled together as one cephalopods group in the present model for the Caribbean ecosystem. Likewise, input Ecopath parameter values on these groups were aggregated. The most important cephalopod species were leatherback, Loligo sp. and Octopus vulgaris. Their annual biomass in the Lesser Antilles pelagic ecosystem ranged from 0.177 t*km-2 to 1.157 t*km-2 (Mohammed et al. 2007b). After aggregating the small and large squid groups into one group, the resulting annual biomass was 1.334 t*km-2. This estimate was close to the maximum range of cephalopod’s annual biomass in the other Ecopath models in the Caribbean areas, which ranged from 0.0051 t*km-2 in Central Atlantic (Vasconcellos and Watson 2004) to 10.0 t*km-2 in Southern Mexican Caribbean (Alvarez-Hernández 2003). In the present model, we used mid-range annual biomass of 5.000 t*km-2 of the similar species (based on the eight ecosystem models in the Caribbean areas) and assumed to be representitave in the Caribbean ecosystem. Its annual average catch density in the region was 0.00027 t*km-2 for local fleets and 0.00003 t*km-2 for foreign fleets (Mohammed et al. 2007a; Sea Around Us database unpublished data from Reg Watson and www.seaaroundus.org). The P/B ratio of cephalopods in the Lesser Antilles pelagic ecosystem model ranged from 4.6 yr-1 to 5.5 yr-1 (Mohammed et al. 2007b), and was then 5.05 yr-1 after the aggregation process. The other estimates of P/B ratios documented in adjacent areas in the Caribbean ranged from 1.15 yr-1 Central Atlantic (Vasconcellos and Watson 2004) to 8.3 yr-1 in Costa Rica (Wolff et al. 1998). A mid-range P/B ratio of 4.725 yr-1 was used in the present model based on the eight Caribbean ecosystem models. This value was assumed to be representative for the cephalopods in the entire Caribbean ecosystem. The Q/B values, ranging from 15.860 yr-1 to 18.330 yr-1, estimated by Mohammed et al. (2007b) were based on the gross food conversion efficiency (P/Q) of the species. While, cephalopods Q/B used values in other Ecopath ecosystem models in adjacent areas in the Caribbean ranged from 2.300 yr-1 to 36.500 yr-1. These ranges were both came from the Central Atlantic ecosystem model (Vasconcellos and Watson 2004). Here, a mid-range Q/B estimate of 19.400 yr-1 from the similar species was used in the present models representing the Caribbean ecosystem. Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  89 Cephalopods dietary contents are composed of small and schooling pelagics, reef fishes, cephalopods, crustaceans and benthos, benthic producers and zooplankton (Opitz 1996).  25. Crustaceans and benthos There was no crustaceans and benthos group in the initial model. We included a trophic group for crustaceans and benthos in the present model because of its importance to the Caribbean fishery. The crustaceans and benthos species included in the present model are shown in Table 1. The most important species of this group were conch (Strombus sp.), crabs, shrimps (Penaeus sp.), oyster (Crassostrea virginica) and spiny lobster (Panulirus argus). The annual biomass of the similar species for this group ranged from 0.050 t*km-2 in Bahia Ascencion, Mexican Caribbean (Vidal and Basurto 2003) and in Colombian Caribbean Sea (Duarte and Garcia 2002) to 23.935 t*km-2 in Southeastern Caribbean (Mohammed 2003a). The extremely high range of annual biomass (842.000 t*km-2) of the similar species in the Southern Mexican Caribbean was not used to factor the mid-range value of annual biomass for the crustaceans and benthos group. Consequently, the mid-range value for the similar specis of crustaceans and benthos annual biomass used in the present model was 11.993 t*km-2, representing the Caribbean ecosystem. Based on the information from Sea Around Us database (unpublished data from Reg Watson and www.seaaroundus.org), the average annual catch density of crustaceans and benthos species was 0.01419 t*km-2 and 0.00132 t*km-2 for local and foreign fleets, respectively. The most commercially important crustacean fisheries in the Caribbean region are spiny lobsters and conch (Mahon 1990; Opitz 1996). There were no catches data reported of this group from LAPE national fisheries report. The P/B estimates of this group ranged from 0.31 yr-1 in Caribbean coral reef ecosystem (Opitz 1996) to 30.0 yr-1 in Costa Rica (Wolff et al. 1998). Here, we used mid-range P/B estimate of 15.155 yr-1, based on the eight Caribbean ecosystem models, for the crustaceans and benthos group. The latter value was assumed to be representative in the Caribbean ecosystem. Other published models in adjacent areas of the Caribbean used Q/B values of similar species ranging from 2.370 yr-1 in Southern Mexican Caribbean (Alvarez-Hernández 2003) to 150.000 yr-1 in Costa Rica (Wolff et al. 1998). In the present model, a mid-range Q/B estimate of 76.185 yr-1 was used (based on the eight Caribbean ecosystem models) representing the whole of the Caribbean ecosystem. Crustaceans and benthos diet information was obtained from Opitz (1996). According to this study, crustaceans and benthos diet contents are composed of reef fishes, cephalopods, crustaceans and benthos, benthic producers, zooplankton, phytoplankton and detritus.  26. Zooplankton Chaetognatha, Copepoda, Euphausiacea, Hydrozoa, Hyperiidae, Mysidacea, Scyphozoa, ichtyoplankton, macroplankton, meroplankton, planktonic decapods, larvae, and fish eggs comprised the zooplankton group. The zooplankton annual biomass in the Lesser Antilles was obtained from the results of hydroacoustic and pelagic trawl surveys by Melvin et al. (2007) conducted under the LAPE project. The annual biomass of large zooplankton in the area was estimated to be 9.636 t*km-2 while small zooplankton was estimated to amount to 40.000 t*km-2 (Mohammed et al. 2007b). Summing these estimates, a total biomass density for zooplankton of 49.636 t*km-2 was obtained. We incorporated the latter value in the present model and assumed to be representative in the Caribbean ecosystem. The P/B ratio estimates, ranging from 8.7 yr-1 to 17.3 yr-1, for zooplankton in Lesser Antilles pelagic ecosystem were initially derived from the Caribbean coral reef ecosystem model (Mohammed et al. 2007b). This P/B range was aggregated resulting to 12.9 yr-1, and was incorporated in the present model representing the Caribbean ecosystem. Food web models and data for the Caribbean model, J.L. Melgo et al.  90 The Q/B estimate of zooplankton in Mohammed et al. (2007b) for Lesser Antilles pelagic ecosystem ranged from 29.000 yr-1 to 57.700 yr-1 and was 43.350 yr-1 after aggregation process. The latter value was used in the present model and assumed to be representative for zooplankton in the Caribbean ecosystem. Diet information of zooplankton group was obtained from Heileman et al. (2007). In their study, they stated that this group feed on crustaceans and benthos (juveniles), zooplankton, benthic producers, phytoplankton and detritus.  27. Benthic producers Benthic producers were composed of benthic autotrophs, marine algae species, and symbiotic algae. There is no benthic prducers group in the initial model. However, we added herein the trophic group for benthic producers because it is an ecologically important component in the Caribbean ecosystem.  The annual biomass of this group ranged from 0.500 t*km-2 in Costa Rica (Wolff et al. 1998) to 27.871 t*km-2 in Southeastern Caribbean (Mohammed 2003a). The extremely high range of annual biomass (1641.0 t*km-2) of the similar group in the Southern Mexican Caribbean was excluded from the average annual biomass value for the crustaceans and benthos group. Consequently, the mid-range value for benthic producers annual biomass of 14.186 t*km-2 was used in the present model representing the Caribbean ecosystem An intermediate or mid-range P/B value of 71.75 yr-1 for benthic producer group was used in the present model for the Caribbean ecosystem. This value was derived from the eight models in the adjacent areas in the Caribbean that ranged from 13.25 yr-1 in Southern Mexican Caribbean (Alvarez-Hernández 2003) to 280.0 yr-1 in Coral reef ecosystem (Opitz 1996).  28. Phytoplankton The phytoplankton input parameters in the Lesser Antilles pelagic ecosystem model by Mohammed et al. (2007b) were taken from primary production study by LAPE project and the production/biomass of phytoplankton by Forget (2007). The annual biomass and P/B ratio of phytoplankton used in the present model were 32.0 t*km-2 and 42.8 yr-1, resepectively. These values were derived from the Lesser Antilles pelagic ecosystem developed by Mohammed et al. (2007b), and assumed to be representative in the Caribbean ecosystem.  29. Detritus Particulate organic materials (POM) and dissolved organic materials (DOM) along the water column or in the bottom of the ocean constituted the detritus group in the model. The detritus annual biomass was estimated from Pauly et al. (1993) empirical equation on detritus biomass as function of primary production and euphotic depth by Rajendra et al. (1991) and from primary production results by Forget (2007). Mohammed et al. (2007b) estimated that the annual biomass for detritus was 15.075 t*km-2 in the Lesser Antilles pelagic ecosystem. This value is extremely low compared to the other annual biomass range of detritus in the Caribbean areas that range from 13.000 t*km-2 in Bahia Ascencion, Mexican Caribbean (Vidal and Basurto 2003) to 600 t*km-2 in Southern Mexican Caribbean (Alvarez-Hernández 2003). We used an intermediate or mid-range annual biomass estimate of 306.500 t*km-2 (based on eigth ecosystem models) for detritus group in the present model representing the Caribbean ecosystem. Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  91  BALANCING THE MODEL The unbalanced model for the Caribbean is shown in Table 19. In order to obtain a balanced solution, two major levels of verification have been made. First, the ecotrophic efficienty (EE) terms were examined to evaluate the balance among our trophic groups and within the whole system. If a particular group was ‘unbalanced’ (with an EE higher than 1.00), this indicated that biomass or P/B values for the groups were underestimated, or that the Q/B was overestimated. Secondly, it was important to make sure that gross efficiency (GE), which is the ratio of production to consumption (P/Q), was always within the 0.1 – 0.3 range. According to Christensen and Pauly (1992), GE should range from 10% to 30%, with exception of top predators, e.g., marine mammals and seabirds, which can have lower GE (between 0.1 and 1%).  Small, fast growing fish larvae or nauplii or bacteria are also exempted because they can have higher GE (between 25 and 50%) (Christensen and Pauly 1992). Table 19. Input data for the Ecopath model of the Caribbean. Unbalanced values are shown in bold.  Ecopath group Trophic level B (t*km-2) P/B (year-1) Q/B (year-1) EE 1 Minke whales 3.98 0.005 0.099 8.421 0.000 2 Fin whales 3.26 0.006 0.099 4.161 0.000 3 Humpback whales 3.84 0.070 0.099 5.081 0.000 4 Brydes whales 3.92 0.004 0.050 6.260 0.000 5 Sei whales 3.26 0.001 0.020 6.178 0.000 6 Blue whales 3.25 0.002 0.060 3.398 0.000 7 Sperm whales 4.45 0.019 0.050 5.030 0.000 8 Killer whales 4.19 0.0003 0.020 9.468 1.724 9 Beaked whales 4.05 0.0001 0.036 9.933 0.000 10 Small cetaceans 3.93 0.005 0.030 14.404 3.350 11 Seabirds 3.55 0.0002 0.130 73.690 0.000 12 Seaturtles 4.31 0.037 0.835 3.535 0.464 13 Large tunas and billfishes 4.34 0.027 1.250 15.530 0.555 14 Small tunas 4.08 0.012 1.960 19.610 8.965 15 Dolphinfish 4.52 0.028 4.720 20.000 0.763 16 Flyingfish 3.33 0.208 4.000 24.760 0.534 17 Other offshore predators 2.06 1.627 1.863 9.270 0.057 18 Pelagic sharks 4.52 0.012 0.400 10.000 0.265 19 Coastal and demersal sharks and rays 3.98 0.200 0.356 5.480 0.181 20 Scombrids 4.31 0.066 1.090 10.310 0.483 21 Small and schooling pelagics 3.29 16.575 2.940 23.650 0.400 22 Reef fishes 3.15 0.654 1.755 21.000 10.004 23 Coastal predators 3.74 1.260 0.720 7.220 1.979 24 Cephalopods 3.45 5.000 4.725 19.400 0.188 25 Crustaceans and benthos 2.38 11.993 15.155 76.185 3.800 26 Zooplankton 2.25 49.636 12.900 43.350 0.669 27 Benthic producers 1.00 14.186 71.750 - 0.222 28 Phytoplankton 1.00 32.000 42.800 - 0.962 28 Detritus 1 306.500 - - 0.688  In the Ecopath approach, estimates of the different parameters (Biomass, P/B, Q/B or diet composition) for all groups of the foodweb can be adjusted to bring the groups and the model into balance. Since there were multiple connections among groups, a change in the estimate for a trophic group may in turn have changed the degree of balance of other groups that connected with it. While these adjustments could be seen as a “trial and error” (or haphazard) process, the systematic method we used here [based on methodology developed by Morissette et al. (in press) and Savenkoff et al. (2007)] Food web models and data for the Caribbean model, J.L. Melgo et al.  92 all allowed to make the changes within a range of possible values based on similar species and foodwebs in the Caribbean area. Doing so, we avoided the arbitrary parameter adjustments that may lead to unnecessary erosion of model realism (Okey 1999). Modifications to the original model needed to reach a balanced solution are listed below: 1. The EE for small cetaceans was higher than 1.000, indicating that there was not enough biomass in the model to account for all sources of mortality on this group. Given the unavoidable degree of uncertainty associated with density estimates derived from the global model used by Kaschner (2004) to provide small cetacean density estimates, we considered a substantial increase of small cetacean biomass density justifiable and likely to still fall within the bounds of uncertainty. To the new estimate was the same as the one used by Vidal and Basurto (2003) in Bahia Ascencion, Mexican Caribbean, where similar species of dolphins reach a biomass of 0.040 t*km-2. 2. Killer whales were an important group of marine mammals in the Caribbean ecosystem, and were also exploited by whaling. In the initial model, their EE was higher than 1.000. This means that their biomass was too low, or biomass of their prey was insufficiently low. Since they partially feed on small cetaceans (which was also unbalanced, see below), their EE was set to 0.950 to determine which biomass would be needed to support the level of exploitation of this group. However, this was done only after having incorporated the above increase the biomass for the small cetaceans group and thus having creating more food for the killer whales. The resulting killer whale biomass (0.000526 t*km-2) was acceptable and within the range of possible values provided by Kaschner’s (2004) database. 3. There was also an EE higher than 1.000 for small tunas. To compensate for that, we first decreased the proportion of small tunas in the diet of large tunas, based on the proportion of the key-species of small tuna (skipjack tuna) in the diet of the key-species of large tunas (yellowfin tuna). The diet of small tunas on small tunas (cannibalism) was also decreased to a minimal proportion of 0.010. The main predator of small tunas in our system was coastal and demersal sharks and rays. The proportion of small tunas in the diet of this group was also reduced, assuming that the coastal and demersal sharks and rays group features, a more coastal diet for our study area, compared to the more pelagic environment used by small tunas. The new proportion of small tunas in the coastal and demersal sharks and rays group was adjusted by increasing the proportions of small and schooling pelagics and other offshore predators in the diet. The biomass, P/B and Q/B of coastal and demersal sharks and rays were also decreased to minimal values (within the range of possible values for surrounding areas) of 0.040 t*km-2 (Duarte and Garcia 2002), 0.112 and 1.800 yr-1 (Vasconcellos and Watson 2004), respectively. Finally, in order to get a balanced solution for the small tunas group, we used maximal values of 2.500 and 25.000 yr-1 for small tunas’ P/B and Q/B, respectively, based on the range of possible values for similar and neighboring ecosystem models. These inputs were the same as the values used by Olson and Watters (2003) in eastern tropical Pacific ecosystem. Biomass was also adjusted using an EE of 0.950. The resulting B (0.0145 t*km-2) was part of the range of possible values, and similar to what was seen in southeastern Caribbean ecosystem (Mohammed 2003a) and Lesser Antilles pelagic ecosystem (Mohammed et al. 2007b). 4. The EE of reef fishes was also higher than 1.000. In order to compensate for that, we used an EE of 0.950 and let Ecopath calculate the biomass required to reach a balanced scenario. The resulting value (24.272 t*km-2) was within the range of possible values, and about half the value used by Opitz (1996) for Caribbean coral reefs. 5. The EE for coastal predators was higher than 1.000, indicating that there was not enough biomass in the model to account for all sources of mortality on this group. The main problem here was a very high cannibalism. Thus, we allowed a proportion of 0.010 instead of 0.140 for cannibalism. The remaining proportion was redistributed between small and schooling pelagics, and crustaceans and benthos, the two main preys of coastal predators. 6. For crustaceans and benthos, cannibalism also seemed overestimated and unrealistic.This was due to aggregation of distinct species in the original model that were now part of the same group. Consequently, we reduced the cannibalism of crustaceans and benthos from 0.238 to Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  93 0.050, and increased the proportion of detritus in the diet (the main prey) to compensate for that change. Finally, we increased the P/B to 20.0 yr-1 (close to the maximal value from Wolff et al. [1998] in Costa Rica), and used EE=0.95 to let Ecopah calculate a biomass (51.201 t*km-2, which was in the range of possible values for this group in the area), to finally reach a balanced solution for crustaceans and benthos group. 7. The changes on crustaceans and benthos created slight imbalance in benthic producers and phytoplankton. Thus, we used an EE of 0.6 for both groups to get a new biomasses of 23.157 t*km-2 (Mohammed 2003a) and 79.081 t*km-2, for benthic producers and phytoplankton groups, respectively. The new biomass estimate of the latter group was off from the possible values, 0.426 t*km-2 (Olson and Watters 2003) to 47.0 t*km-2 (Alvarez-Hernández 2003), found in similar ecosystem models. However, this was not considered to have huge variation compared to what used by Opitz (1996) for phytoplankton annual biomass in the Caribbean coral reef ecosystem, which was 1300 t*km-2.  The latter value was considered as an outlier for our ranges. 8. Changed the diet of zooplankton on detritus from 0.2 to 0.15 (re-distributing the 0.05 to phytoplankton) to balance for detritus. 9. Higher biomass of crustaceans and benthos also created an imbalance of detritus. Consequently, we slightly reduced the proportion of this group (crustaceans and benthos) in the diet of predators (reef fish and crustaceans and benthos) to create less predation, and thus a smaller biomass calculated with the EE of 0.95. The final biomass of crustaceans and benthos is 61.755 t*km-2 (this was still within the range of possible values from adjacent ecosystems) using an EE of 0.95. The new biomass estimate used for crustaceans and benthos group falls a bit outside the range of the possible values, 0.050 t*km-2 (Vidal and Basurto 2003) to 23.935 t*km- 2 (Mohammed 2003a), found in similar ecosystem models. However, this new crustaceans and benthos biomass estimate was not considered to have huge variation compared to what was used in the southern Mexican Caribbean ecosystem model (842 t*km-2) by Alvarez-Hernández (2003), which was considered as an outlier for our ranges. The final balanced model for the Caribbean region is given in Table 20. A diet matrix showing the proportion of each prey in all predators’ diets is provided in Table 21.  Food web models and data for the Caribbean model, J.L. Melgo et al.  94 Table 20. Balanced version of the Ecopath model of the Caribbean. Estimated parameters are shown in bold.  Ecopath group Trophic level B (t*km-2) P/B (year-1) Q/B (year-1) EE 1 Minke whales 3.890 0.005 0.099 8.421 0.000 2 Fin whales 3.220 0.006 0.099 4.161 0.000 3 Humpback whales 3.740 0.070 0.099 5.081 0.000 4 Brydes whales 3.860 0.004 0.050 6.260 0.000 5 Sei whales 3.220 0.001 0.020 6.178 0.000 6 Blue whales 3.200 0.002 0.060 3.398 0.000 7 Sperm whales 4.240 0.019 0.050 5.030 0.000 8 Killer whales 4.070 0.0005 0.020 9.468 0.950 9 Beaked whales 3.880 0.000 0.036 9.933 0.000 10 Small cetaceans 3.800 0.040 0.030 14.404 0.784 11 Seabirds 3.370 0.0002 0.130 73.690 0.000 12 Seaturtles 4.060 0.037 0.835 3.535 0.298 13 Large tunas and billfishes 3.980 0.027 1.250 15.530 0.642 14 Small tunas 3.600 0.014 2.500 25.000 0.950 15 Dolphinfish 4.440 0.028 4.720 20.000 0.741 16 Flyingfish 3.290 0.208 4.000 24.760 0.543 17 Other offshore predators 2.050 1.627 1.954 9.270 0.169 18 Pelagic sharks 4.280 0.012 0.400 10.000 0.265 19 Coastal and demersal sharks and rays 3.740 0.040 0.112 1.800 0.596 20 Scombrids 4.090 0.066 1.090 10.310 0.346 21 Small and schooling pelagics 3.220 16.575 2.940 23.650 0.528 22 Reef fishes 2.950 21.233 1.755 21.000 0.950 23 Coastal predators 3.410 1.260 0.720 7.220 0.675 24 Cephalopods 3.250 5.000 4.725 19.400 0.526 25 Crustaceans and benthos 2.100 51.201 20.000 76.185 0.950 26 Zooplankton 2.200 49.636 12.900 43.350 0.847 27 Benthic producers 1.000 14.186 71.750 - 0.950 28 Phytoplankton 1.000 32.000 100.000 - 0.950 29 Detritus 1 306.5 - - 0.903 Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  95 Table 21. Diet matrix for the Ecopath model of the Caribbean  Prey / predators 1 2 3 4 5 6 7 8 9 10 11 12 13 1 Minke whales 2 Fin whales 3 Humpback whales 4 Brydes whales 5 Sei whales 6 Blue whales 7 Sperm whales 8 Killer whales 9 Beaked whales 10 Small cetaceans        0.150  0.00001 11 Seabirds 12 Seaturtles 13 Large tunas and billfishes             0.008 14 Small tunas             0.02 15 Dolphinfish             0.003 16 Flyingfish             0.011 17 Other offshore predators 0.010  0.009 0.149 0.0007  0.000 0.146 0.178 0.370 0.321  0.191 18 Pelagic sharks 19 Coastal and demersal sharks and rays 20 Scombrids 0.003   0.070         0.008 21 Small and schooling pelagics 0.049 0.001 0.034 0.536 0.0007   0.264  0.042  0.020 0.51 22 Reef fishes           0.108 0.480 23 Coastal predators 0.012 0.0002 0.017  0.0007      0.144 0.450 0.065 24 Cephalopods       0.996 0.365 0.687 0.588   0.13 25 Crustaceans and benthos       0.004  0.089  0.369 26 Zooplankton 0.026 0.0981 0.040 0.245 0.098 0.100  0.075 0.046  0.058  0.054 27 Benthic producers            0.050 28 Phytoplankton 29 Detritus  Import 0.900 0.901 0.900  0.900 0.900  SUM 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 Food web models and data for the Caribbean model, J.L. Melgo et al.  96 Table 21 (cont.). Diet matrix for the Ecopath model of the Caribbean.  Prey / predators 14 15 16 17 18 19 20 21 22 23 24 25 26 27 1 Minke whales 2 Fin whales 3 Humpback whales 4 Brydes whales 5 Sei whales 6 Blue whales 7 Sperm whales 8 Killer whales 9 Beaked whales 10 Small cetaceans     0.001 11 Seabirds 12 Seaturtles     0.07608 0.005 13 Large tunas and billfishes 0.023    0.04505 14 Small tunas 0.01 0.005   0.003 0.03 0.015 15 Dolphinfish 0.001 0.162   0.001 0.003 16 Flyingfish 0.054 0.686     0.063 17 Other offshore predators 0.527 0.001   0.12412 18 Pelagic sharks     0.00801 19 Coastal and demersal sharks and rays      0.01 20 Scombrids 0.003 0.003   0.03403 0.01 21 Small and schooling pelagics 0.169 0.001 0.081  0.02903 0.057 0.003 0.022 0.013 0.208 0.089 22 Reef fishes    0.027  0.562  0.008 0.029 0.033 0.03 0.004 23 Coastal predators 0.001 0.074   0.38338  0.488   0.01 24 Cephalopods 0.173 0.026   0.29029 0.024 0.198  0.004 0.05 0.017 0.002 25 Crustaceans and benthos      0.283  0.115 0.704 0.655 0.662 0.05 0.162 26 Zooplankton 0.039 0.042 0.919  0.001  0.233 0.855 0.062 0.017 0.185 0.029 0.02 27 Benthic producers    0.973  0.001   0.115  0.017 0.185 0.018 28 Phytoplankton         0.073 0.027  0.028 0.65 29 Detritus     0.004 0.015      0.702 0.15  Import  SUM 1.000 1.000 1.000  1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  97   Figure 3. Comparison of the biomass structure (excluding benthic producers and plankton) of the unbalanced vs. balanced Ecopath models The changes in the general trophic structure from the unbalanced to balanced models are presented in Figure 3. In both models, the system was dominated by benthic species and crustaceans, but coastal predators and reef fishes were also important. The group that changed the most as a result of balancing the model, in terms of biomass, was the ‘crustaceans and benthos’ group. This group tripled in terms of biomass density in the balanced scenario, but it was still within the range of possible values for similar species in other models of the Caribbean islands. Food web models and data for the Caribbean model, J.L. Melgo et al.  98  TIME SERIES DATA CPUEs The timeseries catch per unit effort (CPUE) data for all marine species in the Caribbean region were obtained from published papers and reports during a thorough search of aquatic sciences and fisheries abstracts (ASFA), google scholar, web of science and fish and fisheries worldwide online databases. In the Caribbean region, we found a total of 14 sources for fish and shellfishes timeseries CPUE information (Table 22). This included CPUE data sets starting from 1960’s to 2006’s, however, there were time lags with each source for different species (Table 22). The CPUE timeseries data were available to albacore, bigeye tuna, yellowfin tuna, atlantic white and blue marlins, billfishes, dolphinfish, flyingfish, wahoo, sharks, grouper, miscellaneous reef fishes, snapper, Caribbean queen conch and spiny lobster. These species were assigned and aggregated to our established trophic groups with a total of eight trophic groups: large tunas and billfishes, dolphinfish, flyingfish, pelagic sharks, scombrids, coastal predators, reef fishes, and crustaceans and benthos group (Table 22). However, the CPUE information of all trophic groups was not easily compared across studies due to the difference in sampling gears or efforts used in estimating the catch per unit effort (e.g. tonnes per trip, tonnes per hooks, tonnes per dives, tonners per line fisher). Hence, we noted the different CPUEs units used of each species in the different literatures by aggregating similar species CPUE’s units (Tables 23-26). The catch per unit effort (CPUE) for mutton snapper line and pot fisheries in Puerto Rico was described by Cummings (2007). This study covered the updated and recent CPUE information for mutton snapper (Lutjanus analis) in Puerto Rico. It contains CPUE data from 1989 to 2006 for line fishery and for the pot (trap) fishery of mutton snapper. Cummings (2007) also described the procedures used to update the mutton snapper Puerto Rico CPUE indices. The mutton snapper was categorized in our model as belonging to the “coastal predator” trophic group (Table 22). Consequently, its CPUE data were assigned to the other coastal predators, and aggregated with other sources describing species of that group that have similar CPUE units as tonnes per trips (Table 24). The stock assessment of the Caribbean queen conch (Strombus gigas) by Southeast Data, Assessment, and Review (SEDAR 2007) reported CPUE’s information for queen conch from the period of 1981 - 2005 in St. Croix and 1989 - 2005 in Puerto Rico. This species was part of the trophic group “crustaceans and benthos” in our model. The CPUE unit of this group was tonnes per trips (Tables 22 and 23). According to the results of SEDAR (2007), queen conch stocks and CPUEs are declining. Furthermore, this species is an important commercial invertebrate in the region, which is currently being exploited by many conch gatherers in the coastal areas (Boulon and Clavijo 1986; Mahon 1987; Opitz 1996; SEDAR 2007). The primary fishing ground of queen conch in the Caribbean is the U.S. Virgin Islands and its adjacent areas (Opitz 1996). Four-year fishery monitoring of the tiger grouper (Mycteroperca tigris) spawning aggregation in Puerto Rico was investigated by Matos-Caraballo et al. (2006) (Table 22). They reported the decline of tiger grouper’s CPUE from 1996 to 1998 (Table 26). We summarized these data into an aggregated timeseries representing reef fishes with CPUE units being ‘tonnes per boat per day’. The blue shark (Prionace glauca) caught by Venezuelan pelagic longline fishery was assessed by Arocha et al. (2005). Their main focus of the study was the length composition of blue shark from the Venezuelan pelagic longline fishery for the period of 1994 – 2003 (Table 22). According to Arocha et al. (2005), harvest of blue sharks during the first quarter off Guyana and Suriname increase progressively towards Trinidad and further in the Caribbean during the second and third quarter. Fishers move further towards the Caribbean areas and concentrate their harvest in area during the fourth quarter of the year. The blue shark species was categorized in our report as pelagic sharks. We aggregated its CPUE information under ‘pelagic sharks’ with CPUE units being ‘numbers of species per 1000 hooks’ (Table 26). Esquivel-Valle (2005) constructed standardized indices of spiny lobster abundance using the data from NOAA Fisheries Trip Interview Program (TIP) in Puerto Rico. The available CPUE data for spiny lobsters in the area covered a longer time period from 1983 to 2003. According to Esquivel-Valle (2005) the fishery for spiny lobster in the area operates year round. Their study also noted the irregular sampling of the Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  99 catches that may underrepresent the total landings. The spiny lobster CPUE’s was aggregated as crustaceans and benthos group with CPUE units as ‘tonnes per trip’ (Tables 21 and 22). Nemeth (2005) investigated the population characteristics of a recovering U.S. Virgin Islands red hind (Epinephelus guttatus) spawning aggregation. The study presented five years of red hind CPUE, wherein a fluctuation in the population trend of the species was observed. Together with the other reef-associated fishes mentioned above, the timeseries data of this species were summarized and aggregated representing the reef fish group with CPUE units as ‘tonnes per trap’ (Tables 21 and  25). Saul et al. (2005) conducted a preliminary analysis and standardized catch per unit effort indices for yellowtail snapper (Ocyurus chrysurus) fishery resulting independent data from 1988 to 2001 in Puerto Rico (Table 22). Their nominal hook and line CPUE for snappers, which categorized as ‘coastal predators’ herein, increased in the early 199o’s, and remained stable from 1993 to 2001 (Table 26). This species was aggregated representing coastal predators group with CPUE unit as ‘tonnes per hook and line’. Early stock assessment of SEDAR (2005b) dealt with the spiny lobster (Panulirus argus) CPUE data for the 1974-2003 period and yellowtail snapper (Ocyurus chrysurus) for 1984-2003 (Table 22). In this report, they documented a declining catch rates or CPUE’s of spiny lobster and yellowtail snapper in the area. In our analysis, the spiny lobster and snapper were categorized as crustaceans and benthos and coastal predators, respectively. The CPUEs units used by SEDAR (2005b) were ‘tonnes per trip’ (Table 23) for spiny lobsters and ‘tonnes per linefisher’ (Table 26) for yellowsnapper. The Caribbean red snapper (Lutjanus purpureus) and yellowedge grouper (Epinephelus flavolimbatus) biomass assessment of the 1981-2000 periods was done by Mendoza and Larez (2004) (Table 22). The authors showed decreasing CPUEs trends for red snapper and yellow grouper in the Caribbean because of the overexploitation by handliner and longliner operations in the region. They also concluded that red snapper is more exploited than yellow grouper, and they inferred that this is because the differences in habitat depth between species and the greater efficiency of handliners targeting the red snapper (Mendoza and Larez 2004). These species were grouped and aggregated as coastal predators for snapper and reef fishes for yellowedge grouper with CPUE units being ‘tonnes per lines per day’ (Table 24). The study of Marcano et al. (2002) provides CPUE (tonnes per 100hooks) information for the small longline tuna fishery from eastern Venezuela during the period of 1986 to 2000 (Table 22). This information included catch and efforts for tunas, wahoo, billfishes and sharks in the Caribbean and Atlantic subareas. CPUE data of these species were assigned as large tuna and billfishes, scombrids and pelagic shark in our model (Table 25). Marcano et al. (2002) reported decreasing trends of average annual CPUE’s for most species in the Caribbean. Among all the species, yellowfin tuna was the most important species contributing higher percentages of the total catches in the Caribbean and Atlantic (Marcano et al. 2001). Parker (2002) proposed a preliminary analysis for flyingfish fishery of Barbados. Their study recorded total landings, catch rates and fishing efforts from 1985 – 2001 (Table 22). Based on their results, CPUEs of flyingfish in the area was fluctuating over time. Flyingfish CPUE unit was ‘tonnes per trip’ (Table 23). Exploitation trends for demersal reef fishes in Grenada were studied by Jeffrey (2000). The species included in this study were snappers, groupers and miscellenaous reef fishes (i.e. snook, squirrelfishes, parrotfishes, tilefishes, surgeonfishes, grunts, goatfishes and porgies). CPUE data for snapper species were grouped with coastal predators, and groupers and miscellenaous reef fishes were aggregated to the reef fishes group with CPUE units as ‘tonnes per trip’ (Tables 22 and 23). Their study revealed that reef fishery resources in the area are declining, as well as monthly catch per unit effort for groupers and snappers declined with increasing fishing intensity during the study period, suggesting overfishing (Jeffrey 2000). Finally, Mahon and Oxenford (1999) dealt with the precautionary assessment and management of dolphinfish in the Caribbean. Their study contained CPUE information for dolphinfish around the Caribbean waters. Mahon and Oxenford (1999) stated that the continuation of the high amount of catches of the dolphinfish in the area could lead to stock depletion. The dolphinfish CPUE units used in their study were ‘tonnes per trip’ and ‘fish per 100 trips’ (Tables 23 and 26). These catch rates were incorporated into our model analysis. Food web models and data for the Caribbean model, J.L. Melgo et al.  100 Table 22. The different sources of trophic group CPUE information available for the Caribbean region. Source Species studied Ecopath Group Year covered Countries CPUE’s unit Cummings 2007 Mutton Snapper Coastal predators 1989-2006 Puerto Rico tonnes/trip SEDAR 2007 Caribbean queen conch Crustaceans and Benthos 1989-2005 Puerto Rico tonnes/trip SEDAR 2007 Caribbean queen conch Crustaceans and Benthos 1981-2005 St.Croix tonnes/trip Esquivel-Valle 2005 Spiny lobster Crustaceans and Benthos 1984-2003 Puerto Rico tonnes/trip SEDAR 2005a Spiny lobster Crustaceans and Benthos 1974-2003 Virgin Islands tonnes/trip Parker  2002 Flyingfish Flyingfish 1985-2001 Barbados tonnes/trip Jeffrey 2000 Snapper Coastal predators 1986-1993 Grenada tonnes/trip Jeffrey 2000 Grouper Reef fishes 1986-1993 Grenada tonnes/trip Jeffrey 2000 Miscellaneous reef fishes Reef fishes 1986-1993 Grenada tonnes/trip Mahon and Oxenford 1999 Dolphinfish Dolphinfish 1961-1989 Barbados tonnes/trip Mahon and Oxenford 1999 Dolphinfish Dolphinfish 1982-1989 Grenada tonnes/trip Mahon and Oxenford 1999 Dolphinfish Dolphinfish 1978-1989 Puerto Rico tonnes/trip Mahon and Oxenford 1999 Dolphinfish Dolphinfish 1985-1989 St. Lucia tonnes/trip Mendoza and Larez 2004 Southern red snapper      Coastal predators 1981-1999 Southeastern Caribbean tonnes/lines/day Mendoza and Larez 2004 Yellow grouper Reef fishes 1981-1999 Southeastern Caribbean tonnes/lines/day Marcano et al. 2002 Yellowfin tuna Large tunas and billfishes 1986-1999 Caribbean Sea tonnes/100hooks Marcano et al. 2002 Albacora Large tunas and billfishes 1986-2000 Caribbean Sea tonnes/100hooks Marcano et al. 2002 Big eye tuna Large tunas and billfishes 1986-2000 Caribbean Sea tonnes/100hooks Marcano et al. 2002 Billfishes Large tunas and billfishes 1986-2000 Caribbean Sea tonnes/100hooks Marcano et al. 2002 Shark Pelagic sharks 1986-2000 Caribbean Sea tonnes/100hooks Marcano et al. 2002 Wahoo  Scombrids 1986-2000 Caribbean Sea tonnes/100hooks Matos-Caraballo et al. 2006 Tiger grouper Reef fishes 1995-1998 Puerto Rico tonnes/boat-day Arocha et al. 2005 Blue shark Pelagic sharks 1994-2003 Trinidad, the Caribbean numbers/1000 hooks Nemeth 2005 Red hind Reef fishes 1997-2003 US Virgin Islands tonnes/traps Saul et al. 2005 Yellowsnapper Coastal predators 1988-2001 Puerto Rico tonnes/hook and line SEDAR 2005b Yellowsnapper Coastal predators 1984-2003 St.Thomas/St.John tonnes/linesfisher Mahon and Oxenford 1999 Dolphinfish Dolphinfish 1978-1989 St. Vincent Fish/100trips  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  101 Table 23. Catch per unit effort (tonnes/trip) by trophic groups in the Caribbean region. Data were derived from Cummings 2007; SEDAR 2007; Esquivel-Valle 2005; SEDAR 2005b; Parker 2002; Jeffrey 2000, and Mahon and Oxenford 1999. 15 16 22 23 25 Year Dolphinfish Flyingfish  Reef fishes Coastal predators Crustaceans and Benthos 1961 0.010260 1962 0.009470 1963 0.005130 1964 0.016970 1965 0.010260 1966 0.009470 1967 0.010660 1968 0.017760 1969 0.030400 1970 0.026050 1971 0.051824 1972 0.040452 1973 0.024984 1974 0.031294 1975 0.052224 1976 0.045281 1977 0.047401 1978 0.058390 1979 0.044860 1980 0.072170 1981 0.037900    61.688562 1982 0.037675    51.709530 1983 0.047546    24.796385 1984 0.023632    34.782628 1985 0.117004 0.062647   40.226885 1986 0.093794 0.037059 0.113646 0.016322 20.877702 1987 0.075233 0.039706 0.098343 0.028582 32.968786 1988 0.111068 0.058235 0.078968 0.016539 31.152399 1989 0.111908 0.060882 0.026949 0.011829 14.407557 1990  0.027353 0.036907 0.012872 10.324577 1991  0.039706 0.037512 0.012667 104.310557 1992  0.043235 0.009872 0.011142 59.970940 1993  0.031765 0.027032 0.014099 43.399383 1994    0.001424 40.072024 1995   0.043000 0.001546 54.474754 1996  0.015882 0.058000 0.001502 52.358285 1997  0.023824 0.050400 0.001386 47.631743 1998  0.020294 0.028950 0.001598 51.640306 1999  0.036176  0.002219 44.195341 2000  0.044118  0.002037 43.778594 2001  0.033529  0.001944 48.845334 2002    0.001918 47.066394 2003    0.003095 49.712851 2004    0.002076 62.746928 2005    0.001633 68.946033 2006    0.001476 Food web models and data for the Caribbean model, J.L. Melgo et al.  102  Table 24. Catch per unit effort (tonnes/lines*day) by trophic groups in the Caribbean region. Data derived from Mendoza and Larez 2004. 22 23 Year Reef fishes Coastal predators 1981 0.000205 0.01012 1982 0.00018 0.0105 1983 0.000162 0.011413 1984 0.000148 0.00563 1985 0.000143 0.004641 1986 0.000158 0.003957 1987 0.000298 0.007685 1988 0.000157 0.004793 1989 0.000159 1990 0.000159 0.011261 1991 0.000099 1992 0.00013 0.007457 1993 0.000059 0.004793 1994 0.000034 0.003348 1995 0.000088 0.002815 1996 0.000098 0.004489 1997 0.000102 0.001902 1998 0.00011 0.002663 1999  0.002283 2000 0.000049   Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  103 Table 25. Catch per unit effort (tonnes/100 hooks) by trophic groups in the Caribbean region. Data were derived from Marcano et al. 2002. 13 18 20 Year Large tunas and billfishes Pelagic sharks Scombrids 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 0.000430 0.000032 1987 0.000486 0.000045 1988 0.000355 0.000025 1989 0.000401 0.000026 0.000005 1990 0.000324 0.000028 0.000026 1991 0.000720 0.000019 0.000020 1992 0.000688 0.000070 0.000005 1993 0.000654 0.000068 0.000008 1994 0.000675 0.000064 0.000080 1995 0.000529 0.000053 0.000026 1996 0.000433 0.000046 0.000040 1997 0.000233 0.000074 0.000018 1998 0.000474 0.000011 0.000002 1999 0.000657 0.000068 0.000004 2000 0.000561 0.000062 0.000003 2001 2002  Food web models and data for the Caribbean model, J.L. Melgo et al.  104 Table 26. Catch per unit effort of trophic groups that has single data sets for CPUE units: Fish/100 trips, numbers/1000hooks, tonnes/traps, tonnes/day/boat, tonnes/lines fisher and tonnes/hook and line.  Data derived from Matos-Caraballo et al. 2006; Arocha et al. 2005; Nemeth 2005; Saul et al. 2005 and SEDAR 2005a. 15 18 22 22 23 23 Dolphinfish Pelagic sharks Reef fishes Reef fishes Coastal predators Coastal predators Year Fish/100trips numbers/1000 hooks tonnes/traps tonnes/day/boat tonnes/lines fisher tonnes/hook and line 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 31.107000 1979 33.960000 1980 23.389000 1981 20.705000 1982 29.933000 1983     0.424290 1984 16.510000    0.350663 1985 32.114000    0.386697 1986 31.950000    0.192831 1987 16.170000    0.189783 1988 31.610000    0.212644 0.000004 1989 30.100000    0.343982 0.000005 1990     0.288760 0.000003 1991     0.401008 0.000010 1992     0.341976 0.000045 1993     0.420043 0.000017 1994  3.880000   0.400830 0.000030 1995  2.290000  0.043000 0.633578 0.000026 1996  1.410000  0.058000 0.742939 0.000029 1997  3.010000  0.050400 0.674515 0.000028 1998  4.850000  0.028950 0.622669 0.000000 1999  4.250000   0.460215 0.000021 2000  2.870000 0.001576  0.571602 0.000025 2001  4.670000 0.001591  0.704565 0.000014 2002  2.620000 0.004453  0.652160 2003  0.800000 0.001102  0.658344 2004   0.007319 Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  105  FISHING EFFORT Fishing effort of single species is the prefered type of data to use as time series to drive changes in biomass for the dynamic simulations in Ecosim (Christensen et al. 2005). Unfortunately, there was only little information available for timeseries fishing effort for single species in the region (Table 27). The timeseries fishing effort data we found on single species were: yellowtail snapper in St. Thomas/St.John (SEDAR 2005a) and Puerto Rico (Saul et al. 2005), spiny lobsters in Puerto Rico (Esquivel-Valle 2005) and flyingfish in Tobago (Potts et al. 2002). We did not incorporated timeseries fishing effort information from multiple species since it was hard to segregate information of their combined fishing efforts data. Timeseries effort data for multiple species in the region were available in several sources (e.g. Mahon et al. 1994b; Mohammed 2003b; Zeller et al. 2003).  Fishing effort timeseries data from single species were used in our simulation analysis. Similar to the CPUE information, the timeseries fishing effort data for all stocks have different fishing effort units. Hence, we categorized the fishing effort by our trophic groups as well as by effort units. Consequently, we have fishing effort series for three trophic groups with number per trip: large tunas and billfishes, flyingfish and crustaceans and benthos (Table 28), one group with number per line fishers and number per hook and line: coastal predators (Table 29), and one group with number per dives and number per traps: crustaceans and benthos (Table 29). Table 27. The different sources of timeseries fishing effort for the Caribbean region Source Species studied Trophic groups Fishing Effort Units Country Potts et al. 2002 Flyingfish Flyingfish no. of trips Tobago Mahon et al. 1994a Pelagic fishes Large pelagic tunas and billfishes no. of trips Gouyave Grenada Mahon et al. 1994a Pelagic fishes Large pelagic tunas and billfishes no. of trips Melville Grenada Esquivel-Valle 2005 Spiny lobsters Crustaceans and benthos no. of trips Puerto Rico Esquivel-Valle 2005 Spiny lobsters Crustaceans and benthos no. of dive Puerto Rico Esquivel-Valle 2005 Spiny lobsters Crustaceans and benthos no. of fish traps Puerto Rico Esquivel-Valle 2005 Spiny lobsters Crustaceans and benthos no. of lobster traps Puerto Rico Saul et al. 2005 Yellowsnapper Coastal predators no. of hook&line Puerto Rico SEDAR 2005a Yellowsnapper Coastal predators no. line fishers St. Thomas/St John  Food web models and data for the Caribbean model, J.L. Melgo et al.  106 Table 28. Fishing effort as ‘number of trips’, for three trophic groups. Data derived from Potts (2002) and Mahon et al. (1994b). 13 16 25 Year Large pelagic tuna and billfishes Flyingfish Crustaceans and benthos 1981 327 1982 314 1983 301 1984 287  144 1985 346  66 1986 387  52 1987 397 1340 82 1988 300 1371 8 1989 299 1644 126 1990 233 1192 124 1991   181 1992  1192 183 1993  1605 146 1994  990 63 1995  1091 133 1996   112 1997  686 61 1998  319 116 1999   151 2000   124 2001   157 2002   104 2003   135  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  107 Table 29. Fishing efforts: ‘no. of line fishers’, ‘no. of hook & line’, ‘no. of dives’, ‘no. of lobster trap’, for two trophic groups. Data derived from Esquivel-Valle (2005); Saul et al. (2005) and SEDAR (2005a). 23 25 Coastal predators Crustaceans and benthos Year no. line fishers no. of hook and line no. of dive no. of fish traps no. of lobster traps 1983 25 1984 25   144 1985 25  1 65 1986 25  3 49 1987 25  20 58 4 1988 25 127.8 0 8 1989 20 680 79 32 1990 28 81 110 8 1991 28 97.2 130 31 7 1992 28 40.5 138 25 4 1993 28 81 118 16 4 1994 28 72.27 41 20 1 1995 25 102.99 119 13 1996 20 60.3 88 17 1997 20 15 45 13 3 1998 20 0 92 10 9 1999 30 43.59 114 25 5 2000 30 45 104 11 5 2001 30 29.49 136 13 5 2002 30  84 14 1 2003 30  108 17  UNCERTAINTY ANALYSES Given the high level of uncertainty in data, parameterization and model structure (Plagányi and Butterworth 2004; Essington 2006; Plagányi et al. 2007), we conducted several levels of uncertainty analyses. Our efforts focused on facilitating data collection that would shed light on the most appropriate choice of model form with which to represent feeding behaviour. Three levels of uncertainty analyses were performed here. First, a sensitivity routine included in Ecopath was used to explore the effects of uncertainty of inputs values on the model’s outputs. A second uncertainty analysis was performed using Ecoranger, a resampling routine based on input probability distributions. Finally, the robustness of our models’ structure was tested with Ecosim by comparing predicted biomasses with time series of observed data. Sensitivity analysis A sensitivity routine was included in Ecopath to allow users to explore the effects of uncertainty on the model results. The method was quite simple, and consisted of plotting relative output changes against relative changes in the inputs. The routine varies all basic input parameters (biomass [B], production to biomass ratio [P/B], consumption to biomass ratio [Q/B], ecotrophic efficiency [EE]) in steps from -50.0% to +50.0% for each trophic group of the model, and then checks what effect each of these steps has for each of the input parameters on all of the “missing” basic parameters for each group in the system (Christensen et al. 2005). The output is then given as the proportion of the difference between the estimated and original parameter to the original parameter, and converted to a percentage (Christensen et al. 2005). This method only re-estimates the parameters for which no data were available, and that were left to be estimated by the model, using the mass-balance constraints. We conducted a sensitivity analysis for biomass, P/B, Q/B and EE input parameters. Our results suggested that the sensitivity of these estimated parameters to a change in input values was relatively low (Appendix 1). A 50.0% change in any of the input parameters of any trophic group generated an overall response of Food web models and data for the Caribbean model, J.L. Melgo et al.  108 about 39% in the estimated parameters of other groups. All the changes in inputs parameters produced by the sensitivity analysis routine had their effect on the EE of other trophic groups. The most impacted factor in our model seems to be the EE. For biomass inputs, a change would affect all and any parameter with an average 24.7% of change, and EE would change by about 26.0% after a change of 50% in any biomass input. For the P/B ratio, a 50.0% change of any group would generate an average response of 73.6% of any parameter, and a response of 63.5% in the estimated EE of other groups. Similarly, for Q/B input values, a 50.0% change of this value for any group would generate an overall response of 21.7% on any parameter, and an average 19.8% change in the estimated EE of other trophic groups. Overall, our sensitivity analysis suggested that potential errors in model results were approximately linearly related to potential error in model parameters, etc. This result was consistent with those of Essington (2006).  This underscores the importance of enhancing the quality of data included in our model. ‘Ecoranger’ analysis To account for the inherent uncertainty of input parameters, a resampling routine called Ecoranger was included in the EwE software and accepted input probability distributions for the biomasses, consumption and production rates, ecotrophic efficiencies, catch rates, and diet compositions. Ecoranger then draws random input variables using the range of possible values for each parameter, and the resulting model was then evaluated (based on least sum of squared residuals and comparison with independent data and physiological and mass-balance constraints) (Christensen et al. 2005). Starting with the initial model and these setups, 10,000 models were run by Ecoranger, until 200 model runs passed the selection criteria, and the best fitting model for the Caribbean islands was used for further analysis. Fitting the model to time series data The Ecosim model behaviour is based on a ‘foraging arena’ theory (Walters and Kitchell 2004), which assumed that predator and prey behaviors cause partitioning of prey populations, which were either available or unavailable to predators. There was continuous change between these two stages for any given potential prey, whether it was hiding from predation in some refuge, or it was out to feed. This availability of prey to predators was called ‘vulnerability’ in Ecosim. Mackinson et al. (2003) demonstrated the importance of setting the vulnerabilities to fit model predictions to time-series data, as Ecosim predictions are very sensitive to this parameter. Using default values for v has strong implications for assumptions about species abundance relative to their carrying capacity (V. Christensen, Fisheries Centre, UBC, personal communication). Basically, it assumed that each group can at most increase the predation mortality they imposed on their prey with a factor of 2.0 (the default v value). A lower value implies a donor driven density-dependant interaction. On the other hand, a higher value involved a predator driven density-independent interaction, in which predation mortality was proportional to the product of prey and predator abundance (i.e., Lotka-Volterra). This implies a high flux rate for prey species in and out of vulnerable biomass pools (Ainsworth 2006). Vulnerabilities were thus adjusted based on the specific ecology of each species or trophic groups (if their behaviour, niche, or diet make them more or less vulnerable to predators). Using the few time series of biomass available for the trophic groups in our model, we compared Ecosim’s projections with observed data, and adjusted v’s and other input parameters (within their range of uncertainty) until we obtained a model configuration that allowed us to reproduce as much as possible the trends in biomass. Using credible models that can reproduce observed historical response to disturbances such as fishing was a useful approach to validating our model in light of the highly uncertain data included in the model. Fitting time series data to model predictions therefore enhances our confidence about the possible impact of removing marine mammals in the ecosystem (Morissette 2007).  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  109 DISCUSSION The present sets of data were used to construct our model for analyzing the trophic interaction of marine mammals and fisheries in the Caribbean ecosystem. This model was revised after the validation made by the local experts on fisheries, species ecology and biology, and in ecosystem modelling in the Caribbean during a workshop held in Barbados (“Whales and Fish interactions: Are great whales a threat to fisheries”, see http://www.lenfestocean.org/whales_fisheries.html) in September 2008. Suggestions made by local experts included the addition of more important commercial fish groups (e.g. billfishes, dolphinfish, flyingfish, pelagic sharks) in our model and better incorporation of local fisheries data from the region (i.e. LAPE national report). These suggestions were followed and we revised our initial model to its actual form. Uncertainty in the input data Several sources of uncertainty influenced the modelling results presented in this report. The most reliable published fisheries catch estimates for commercial fish and non-fishes species in the Caribbean were from Lesser Antilles Pelagic Ecosystem (LAPE) fisheries report and Sea Around Us Project (SAUP). The LAPE database contained recent catches of the Lesser Antilles countries that were obtained from local fisheries departments for the 2001-2005 periods. On the other hand, the Sea Around Us database contained longer time series catch estimates (starting in the 1950s; we used data for the 1987-2004 period) for fish and shellfish groups in the region. However, recent work on catch data is not included in the SAUP database (E. Mohammed, Fisheries Division, Government of Trinidad and Tobago, personal communication). Due to the expected discrepancies between the two data sets, catch estimates of Lesser Antilles countries from LAPE and SAUP databases were compared. This analysis showed that catches in both databases were relatively similar in terms of the total catches for all local countries combined. They only differ in minute decimal places of their total catches. However, if we considered catch data for individual countries, high discrepancies were observed. These differences could be due to the way catches of species were re- allocated to their geographic origin for each country by the SAUP database versus the reported landings in the LAPE database. In addition to this uncertainty in the databases, catch estimates could involve underestimated catches due to unreported landings from recreational fleets and small-scale fisheries (Mahon and Oxenford 1999, Mohammed et al. 2007a) and in species identification of the catch. The catch per unit effort (CPUE) estimates extracted from literatures (e.g. Marcano et al. 2002; Parker et al. 2002; Restrepo et al. 2003; Arocha et al. 2005; Esquivel-Valle 2005; Cummings 2007) were generally based on fleets and landings. Such processes could cause variation and uncertainty in the data because the main objective in real time fishing was to gather as much fish as possible with no regards to species or type; and the used of advance fishing gear and equipment that enable the fishermen to find the most abundant fishing sites, could results to higher catch rates (CPUE). In addition, the scarcity of information for catch rates of single species could be explained by the fact that most fisheries or fishermen follow a generalistic harvesting strategy i.e. fleets harvesting demersal species can fish for large pelagics while on route to their destination (Mahon and Oxenford 1999). The Ecopath parameter estimates (B, P/B, Q/B) of similar species across local studies can vary. Additionally, some of these local studies based their Ecopath parameter values on the same source that we included in our model, for example P/B value for seaturtles used by Olson and Watters 2003, and Mohammed et al. 2007b was derived from similar species of seaturtles from Opitz 1996. Hence, in order to avoid over-representation of some inputs when averaging all values from other models, we used the mid-range (rather than mean) estimates of annual biomass, P/B and Q/B for some trophic groups in the present model. However, it is important to note that the input parameters for most of the trophic groups of our model are based on field surveys, stomach contents analysis and stock assessments in the area. While these estimates from local studies of our trophic groups were relatively reliable, uncertainty still occurs. An example of such uncertainty was the selectivity variations that were linked to field sampling for species abundance. This variation is not usually quantified (Mohammed et al. 2007b). In addition, the mortality rates (and thus their P/B) of large pelagic species (e.g. yellowfin tuna, albacore, billfishes, wahoo) depend on the quality of the available catch data from regional stock assessment (Mohammed et al. 2007b). This stock assessment information includes areas outside the Caribbean ecosystem (Mohammed et al. 2007b) which, therefore, could overestimate the fishing mortality of fish species due to migration out of the study area. Fishing mortality could be underestimated as well, i.e under reporting catches (particularly for dolphinfish), small sized fish species utilized as baits, or discards (Mahon and Oxenford 1999; Mohammed et al. 2007b). As for marine mammal biomass, the updated global biomass Food web models and data for the Caribbean model, J.L. Melgo et al.  110 densities estimates of Kaschner (2004) are so far the most reliable source. The latter developed a model to capture pattern of species occurrence and densities even for areas that have not been covered by marine mammal surveys or for species that are extremely difficult to observe in the wild. Whenever possible, we validated estimates derived from the global, model were validated using abundance estimates from dedicated surveys conducted in similar habitats (Morisette et al. submitted). These standardized abundance estimates of marine mammals are continuously updated with recent published population studies up and compiled in a global marine mammal database by Kaschner (2004). The diet information is important for investigating the interaction of predator-prey relationship (Morissette 2005). By having quantitative diet estimates of cetaceans, fish groups and non-fish groups, we can determined their dynamics with other compartments of the ecosystem, whether cetaceans were competing with fisheries or eat other species that were less important for fishing. In the Caribbean region, the quantitative information on diet composition for cetacean was limited. The only available cetacean diet information reported for our study area specifically was for the Bryde’s whales (Tershy 1992; Heileman et al. 2007), sperm whales (Kawakami 1980; Clarke et al. 1980; Pascoe et al. 1990; Smith and Whitehead 2000; dos Santos and Haimovici 2001; Hickmott 2005), killer whales (Alonso et al. 1999; Heileman et al. 2007), and beaked whales (Debrot and Barros 1992; Hickmott 2005). The majority of stomach contents analyses available were obtained from whale stranding samples. This will introduce biases, however, since samples may not be representative for the actual diet of the whole population. Cetacean diet composition varies over time as well as locations; and thus, what they eat is certainly not a function of what food items are present in the area in the certain quantities and proportions (Lavigne 1996; Holt 2006). We were, however, unable to find diet information for minke whales, fin whales, humpback whales, sei whales and blue whales in the Caribbean areas. According to Mohammed et al. (2007b), most of these large whales are migrating species that seasonally occur in the Caribbean areas, and hence, less studied and observed in the region. Nevertheless, we incorporated the best available quantitative diet information for cetaceans in the area and from the similar ecosystem providing its general diet compositions and the amount of energy contributed by each prey to cetaceans’ diet. With the available literature on the diet and biology of cetaceans (e.g. Pauly et al. 1998b; Perrin et al. 2002), and general knowledge on the biology and ecology of whales, the prey species for cetacean diets were aggregated according to our trophic groups in our model. Strengths and weaknesses of the model Significant effort was devoted to review all the existing Ecopath models available for the region and any available stock assessment studies on commercial fish species. Our model covered the Caribbean region and included many important marine mammal species as well as important commercial fish and non-fish group, in particular fish species harvested by humans (e.g. tuna, billfishes, flyingfish, scombrids, reef fishes) in the Caribbean ecosystem. To address uncertainty, we conducted a sensitivity and Ecoranger analyses on our data input. In this context, the inclusion of data sets from the existing Ecopath models in adjacent Caribbean areas for confidence interval limits of our input data provided ranges of Ecopath parameters values of our trophic groups important in the Caribbean ecosystem. This is an important step in the way we used and analysed Ecopath models, and have been done only few times so far (Morissette 2005; Morissette et al. submitted; Savenkoff et al. 2007; Bundy et al. in press) The inclusion of SAUP data also allowed us to conduct Ecosim analysis and validation over a longer time period, from 1987 to 2005. With this analysis, our model becomes a useful and reliable tool for policy makers in implementing management scenarios appropriate with the available resources.  These efforts made our model more advantageous compared to the rest of Ecopath models in the adjacent Caribbean areas. On the other hand, the lack of local timeseries biomass of marine mammals and other trophic groups, scarcity of timeseries CPUE data, fishing efforts and diet information of cetaceans in the region are serious caveats, although the uncertainty analysis indicated that our major findings are unlikely to change with improved or updated data. Nevertheless, regional research efforts aiming to fill the gaps in data would be useful. Such data would be essential to further test more realistic scenarios between the interaction of marine mammals and fisheries in the Caribbean ecosystem. Further research on marine mammals abundance, catches and diets, as well as fisheries catches in the area would enhance the quality and precision of the model.  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  111 ACKNOWLEDGEMENTS The authors were thankful for the support of Lenfest Ocean Program and for all the participants of the Caribbean workshop in Barbados. Their inputs and comments on the methods and data used for the construction of this model were priceless. Special thanks were also given to Paul Fanning, Robin Mahon, Elizabeth Mohammed and Silvia Opitz for their constructive insights and support in constructing this preliminary version of the model, and to Margaret Columbe and Allison Hoynes-O’Connor for editing the manuscript. Reg Watson and Dirk Zeller were acknowledged for their willingness and assistance in clarifying the Caribbean fisheries data.  REFERENCES Ainsworth, C.H. 2006. Strategic marine ecosystem restoration in Northern British Columbia. University of British Columbia, Canada. Allen, K. 1971. Relation between production and biomass. Fish. Res. Board Can. 28:1573–1581. Alonso, M.K., Pedraza, S.N., Schiavini, A.C.M., Goodall, R.N.P., and Crespo, E.A. 1999. Stomach contents of false killer whales (Pseudorca crassidens) stranded on the coasts of the Strait of Magellan, Tierra del Fuego. Mar. Mamm. Sci. 15:712-724. Alvarez-Hernández, J. H. 2003. Trophic model of a fringing coral reef in the southern Mexican Caribbean [modelo trófico para un arrecife de coral de tipo borde-barrera en el sur del caribe mexicano]. pp 227-335. In D. Zeller, Booth, S., Mohammed, E., and Pauly, D., (Eds). From Mexico to Brazil: Central Atlantic fisheries catch trends and ecosystem models. Fish. Centre Res. Rep. 11(6):227-335. The Fisheries Centre, University of British Columbia, Vancouver, B.C., Canada. Arocha, F.T.,R. , Silva, J., and Marcano, L. 2005. Blue shark (Prionace glauca) length composition from the Venezuelan pelagic longline fleet in the northwestern Atlantic: period 1994-2003. Int. Comm. Conserv. Atl. Tunas, Collect. Vol. Sci. Pap. 53:942-950. Armstrong, A. J. and W. R. Siegfried. 1991. Consumption of Antarctic krill by minke whales. Antarctic Science 3:13-18. Bacci, M.E. 1998. Marine protected areas in the Eastern Caribbean. Caribbean Natural Resources Institute (CANARI) Technical Report 251. Caracas, Venezuela. Barlow, J. 2003a. Cetacean abundance in Hawaiian waters during summer/fall 2002. Administration report, Admin. Rept. LJ-03-13. Southwest Fisheries Science Center (SWFSC), National Marine Fisheries Service (NMFS), National Oceanic and Atmospheric Administration (NOAA), La Jolla, CA, USA. Barlow, J. 2003b. Preliminary estimates of the abundance of cetaceans along the U.S. west coast: 1991-2001. Administrative Report, Admin. Rept. LJ-03-03. Southwest Fisheries Science Center (SWFSC), National Marine Fisheries Service, (NMFS), National Oceanic and Atmospheric Administration (NOAA), La Jolla, CA, USA. Barlow, J. 2006. Cetacean abundance in Hawaiian waters estimated from a summer/fall survey in 2002. Marine Mammal Science, 22: 446-464. Baum, J.K., Myers, R.A., Kehler, D.G., Worm, B., Harley, S.J., and Doherty, P.A. 2003. Collapse and conservation of shark populations in the Northwest Atlantic. Science 299:389-392. Bax, N. J. 1998. The significance and prediction of predation in marine fisheries. ICES J. Mar. Sci. 55:997-1030. Bell, C.D., Solomon, J.L., Blumenthal, J.M., Austin, T.J., Ebanks-Petrie, G., Broderick, A.C., and Godley, B.J. 2007. Monitoring and conservation of critically reduced marine turtle nesting populations: lessons from the Cayman Islands. Anim. Conserv. 10:39-47. Bentivoglio, A.A. 1988. Investigations into the growth, maturity, mortality rates and occurrence of the dolphin (Coryphaena hippurus, Linnaeus) in the Gulf of Mexico. University College of North Wales, Bangor, UK. Master Thesis. pp 37. Best, P.B., Sekiguchi, K., Rakotonirina, B., and Rossouw, A. 1996. The distribution and abundance of humpback whales off southern Madagascar, August-September 1994. Reports of the International Whaling Commission, 46: 323. Blanco, C. and Raga, J.A. 2000. Cephalopod prey of two Ziphius cavirostris (Cetacea) stranded on the western Mediterranean coast. J. Mar. Biol. Assoc. U.K. 80:381-382. Boulon, R.H.J. and Clavijo, I.E. 1986. Utilization of the Virgin Islands biosphere reserve by artisanal fishermen. Virgin Islands Biosphere Reserve Research Report Virgin Islands Resource Managment Cooperative, Virgin Islands National Park. pp 36. Brown, S.G. and Lockyer, C.H. 1984. Whales. pp 717-781. In R. M. Laws, (Ed). Antarctic Ecology. Academic Press, London, UK. Bundy, A., Heymans, S.J.J., Morissette, L., and Savenkoff, C. in press. Seals, cod and forage fish: a comparative exploration of variations in the theme of stock collapse and ecosystem change in Northwest Atlantic ecosystems. Prog. Oceanogr. Calambokidis, J. and Barlow, J. 2004. Abundance of blue and humpback whales in the eastern North Pacific estimated by capture- recapture and line-transect methods. Marine Mammal Science 20: 63-85. CANARI. 1999. Community-base tourism in the Caribbean. Caribbean Natural Resources Institute. pp 22. Chan, C. and Shing, A. 1999. Shark fisheries in the Caribbean: the status of their management including issues of concern in Trinidad and Tobago, Guyana and Dominica. In R. Shotton, (Ed). Case studies of the management of elasmobranch fisheries. Part 1. Food and Agriculture Organization of the United Nations, Rome, Itally. pp 479 Christensen, I., Haug, T., and Øien, N. 1992. A review of feeding and reproduction in large baleen whales (Mysticeti) and sperm whales Physeter macrocephalus in Norwegian and adjacent waters. Fauna Norv. Ser. A 13:39-48. Christensen, V. and Pauly, D. 1992. Ecopath II - a software for balancing steady-state models and calculating network characteristics. Ecol. Model. 61:169-185. Christensen, V. and Walters, C.J. 2004. Ecopath with Ecosim: methods, capabilities and limitations. Ecol. Model. 172:109-139. Christensen, V., Walters, C.J., and Pauly, D. 2005. Ecopath with Ecosim: a user's guide. Fisheries Centre, University of British Columbia, Vancouver, Canada and International Center for Living Aquatic Resources Management, Penang,  Malaysia. pp 131 Clapham, P.J. 2002. Humpback whale - Megaptera novaeangliae. pp 589-592. In W.F. Perrin, Würsig, B., and Thewissen, J.G.M., (Eds). Encyclopedia of Marine Mammals. Academic Press, San Diego, CA. Clarke, M.R., MacLeod, N., Castello, H.P., and Pinedo, M.C. 1980. Cephalopod remains from the stomach of a sperm whale stranded at Rio Grande do Sul in Brazil. Mar. Biol. 59:235-239. Food web models and data for the Caribbean model, J.L. Melgo et al.  112 Clarke, W.G. 1982. Historical rates of recruitment to Southern Hemisphere fin whale stocks. Rep. Int. Whal. Commn 32:305-324. CLME. 2007. Caribbean Large Marine Ecosystem (CLME) Project. Available online at http://cavehill.uwi.edu/cermes/clmeInfo.html, accessed on November 18, 2008. Constable, A. and Gales, N. 2002. Evaluating the strength of interactions between predators and prey species in food webs: an Index of Predator Influence (IPI) on controlling available production of prey. Paper presented during the Intersessional Workshop on Modelling Cetacean-Fishery Interactions, La Jolla. Cummings, N. 2007. Updated commercial catch per unit of effort indices for mutton snapper line and pot fisheries in Puerto Rico, 1983-2006, SEDAR14 AW01. SEDAR-14 Assessment Workshop Report 1. Cummings, N. and Matos-Caraballo, D. 2007. Information on commercial removals of the mutton snapper, Lutjanus analis, in Puerto Rico from 1983 through 2005 and trends in nominal catch per unit effort, SEDAR14-DW-7. SEDAR-14 Data Workshop Report 4. Davis, R.W. and Fargion, G.S. 1996. Distribution and Abundance of Marine Mammals in North-Central and Western Gulf of Mexico: Final Report. Volume II: Technical Report. Technical Report, OCS Study MMS 96-0027. US Department of the Interior, Minerals Management Service. Davis, R.W., Evans, W.E., and Wursig, B. 2000. Cetaceans, Sea Turtles and Seabirds in the Northern Gulf of Mexico: Distribution, Abundance and Habitat Associations. Volume II: Technical Report. Technical Report, OCS Study MMS 2000-003 - USGS/BRD/CR-1999-0005. Texas A&M University at Galveston and the National Marine Fisheries Service. de la Mare, W. K. 1985. On the estimation of mortality rates from whale age data, with particular reference to minke whales (Balaenoptera acutorostrata) in the Southern Hemisphere. Rep. Int. Whal. Commn. 35:239-250. Debrot, A. and Barros, N.B. 1992. Notes on a Gervais' beaked whale Mesoplodon europaeus and a dwarf sperm whale Kogia simus stranded in Curacao, Netherlands Antilles. Mar. Mamm. Sci. 8:172-178. Debrot, A.O., de Meyer, J.A., and Dezentje, P.J.E. 1998. Additional records and a review of the cetacean fauna of the Leeward Dutch Antilles. Caribb. J. Sci. 34:204-210. Die, D. 2004. Status and assessment of large pelagic resources. In R. Mahon, and McConney, P.A., (Eds). Management of large pelagic fisheries in CARICOM countries. FAO Fish. Tech. Pap., Rome. dos Santos, R. A. and Haimovici, M. 2001. Cephalopods in the diet of marine mammals stranded or incidentally caught along southeastern and southern Brazil (21-34degreeS). Fish. Res. (Amsterdam) 52:99-112. Duarte, L.O. and Garcia, C.B. 2002. Testing responses of a tropical shelf ecosystem to fisheries management strategies: a small-scale fishery from the Columbian Caribbean Sea. Fish. Centre Res. Rep. 10:142-149. The Fisheries Centre, University of British Columbia, Vancouver, B.C., Canada. Eckert, K.L. and Eckert, S.A. 1990. Leatherback Sea Turtles in Grenada, West Indies: A Survey of Nesting Beaches and Socioeconomic Status. Report to the Foundation for Field Research and Fisheries Department, Ministry of Agriculture, Lands, Forestry and Fisheries, St George’s, Grenada. Esquivel-Valle, M. 2005. Standardized catch rates of spiny lobsters (Panulirus argus) estimated from the Puerto Rico commercial trip interview program (1980-2003) in SEDAR. In SEDAR 8 Stock Assessment Report. Caribbean Spiny Lobsters and Yellowtail Snapper. Southeast Fisheries Science Center, NOAA, Charleston, SC. Essington, T.E. 2006. Pelagic Ecosystem Response to a Century of Commercial Fishing and Whaling. pp 38-49. In J.A. Estes, DeMaster, D.P., Doak, D.F., Williams, T.M., and L., R.L., Brownell Jr. (Eds). Whales, Whaling, and Ocean Ecosystems. University of California Press, Berkeley, California, USA. Estes, J.A., DeMaster, D.P., Doak, D.F., Williams, T.M., and Brownell Jr., R.L. 2007. Whales, whaling, and ocean ecosystem. University of California Press. Evans, P.G.H. 1998. Biology of cetaceans of the north-east Atlantic (in relation to seismic energy) in M. L. Tasker, and Weir, C., editors, London. pp 418 Finlay, J. 1984. National report for Grenada. pp 184–196. In P.R. Bacon, Berry, F., Bjorndal, K., Hirth, H., Ogren, L., and Weber, M., (Eds). Proceedings of the Western Atlantic Turtle Symposium. University of Miami Press, Miami, USA., San Jose, Costa Rica. FishBase. Available at world wide web http://www.fishbase.org/, accessed on May 2, 2008. Fonteneau, A. and Marcille, J. 1993. Resources, fishing and biology of the tropical tunas of the Eastern Central Atlantic. FAO, Rome. Forget, M.H. 2007. Scientific basis for ecosystem-based management in the Lesser Antilles including interactions with marine mammals and other top predators: phytoplankton community and primary production in the Caribbean waters: the biological oceanography component of the LAPE project. FI:GCP/RLA/140/JPN. FAO Technical Document Barbados. Garrison, L.P., Swartz, S., Martinez, A., Burks, C., and Stamates, J. 2003. A marine mammal assessment survey of the Southeast US Continental Shelf: February - April 2002. NOAA Technical Memorandum, NMFS-SEFSC-492. U.S. Department of Commerce. Gobert, B. 2000. Comparative assessment of multispecies reef fish resources in the Lesser Antilles. Fish. Res. 44:247-260. Goetz, S.J., Prince, S.D., Goward, S.N., Thawley, M.M., and Small, J. 1999. Satellite remote sensing of primary production: an improved production efficiency modeling approach. Ecol. Model. 122:239-255. Grazette, S., Horrocks, J.A., Phillip, P.E., and Isaac, C.J. 2007. An assessment of the marine turtle fishery in Grenada, West Indies. Oryx 41:330-336. Guénette, S., Heymans, S.J.J., Christensen, V., and Trites, A.W. 2006. Food web models and data for studying fisheries and environmental impacts in Eastern Pacific ecosystems. Can. J. Fish. Aquat. Sci. 63:2495-2517. Haug, T., Gjosaeter, H., Lindstrøm, U., and Nilssen, K. T. 1995. Diet and food availability for Northeast Atlantic minke whales (Balaenoptera acutorostrata), during the summer of 1992. ICES J. Mar. Sci. 52:77-86. Haug, T., Lindstrøm, U., Nilseen, K.T., Røttingen, I., and Skaug, H.J. 1996. Diet and food availability for northeast atlantic minke whales, Balaenoptera acutorostrata. Rep. Int. Whal. Commn. 46:371-382. Hawley, C. 1999. A Caribbean whaler speaks up for tradition despite activists’ outcry. The Chicago Tribune, May 25, 1999. Heileman-Manickchand, S.C. 1992. A preliminary survey of clupeoid fishes in Trinidad, West Indies, and their use in live bait fishing. Proc. Gulf  Caribb. Fish. 41:403-415. Heileman, S. 2007. Thematic report for the insular Caribbean sug-region. Caribbean Large Marine Ecosystem (CLME) Project. University of the West Indies, Barbados. pp 63. Heileman, S., Mohammed, E., and Fanning, P. 2007. Scientific basis for ecosystem-based management in the Lesser Antilles including interactions with marine mammals and other top predators: derivation of diet compositions in the Lesser Antilles Pelagic Ecosystem. FAO, Barbados. pp vii+77 Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  113 Hewitt, R.P. and Lipsky, J.D. 2002. Krill. pp 676-684. In Perrin, W.F. Würsig, B., and Thewissen, J.G.M., (Eds). Encyclopedia of Marine Mammals. Academic Press, San Diego, CA. Heymans, S.J.J. 2005. Ecosystem model of the eastern Aleutian and central Gulf of Alaska in 1963. pp 80-105. In S. Guénette, and Christensen, V., (Eds). Food web models and data for studying fisheries and environmental impacts on Eastern Pacific ecosystems. Fish. Centre Res. Rep. 13(1):80-105. The Fisheries Centre, University of British Columbia, Vancouver, B.C., Canada. Hickmott, L.S. 2005. Diving behaviour and foraging ecology of blainville's and cuvier's beaked whales in the northern Bahamas. Department of Marine and Fisheries Biology. University of Aberdeen, Scotland, U.K. Master thesis. pp ii+102 Hjort, J. and Ruud, J.T. 1929. Whaling and fishing in the North Atlantic. Rapp. P.-v. Reun. Cons. perm. int. Explor. Mer. 56:1-123. Holt, S.J. 2006. Whales competing? an analysis of the claim that some whales eat so much that they threaten fisheries and the survival of other whales. Paciano, Italy. pp 86, Horwood, J. 1990. Biology and exploitation of the minke whale. CRC Press, Boca Raton, FL, USA. ICCAT. 2004. Atlantic yellowfin tuna stock assessment session. Col. Vol. Sci. Pap.  56:443-527. Innes, S.D., Lavigne, M., Earle, W.M., and Kovacs, K.M. 1987. Feeding rates of seals and whales. J. Anim. Ecol. 56:115-130. Jackson, J.B.C., Kirby, M.X., Berger, W.H., Bjorndal, K.A., Botsford, L.W., Bourque, B.J., Bradbury, R.H., Cooke, R., Erlandson, J., Estes, J.A., Hughes, T.P., Kidwell, S., Lange, C. B., Lenihan, H.S., Pandolfi, J.M., Peterson, C.H., Steneck, R.S., Tegner, M.J., and Warner, R.R. 2001. Historical overfishing and the recent collapse of coastal ecosystems. Science 293:629-637. Jann, B., Allen, J., Carrillo, M., Hanquet, S., Katona, S.K., Martin, A.R., Reeves, R.R., Seton, R., Stevick, P.T., and Wenzel, F.W. 2003. Migration of a humpback whale (Megaptera novaeangliae) between the Cape Verde Islands and Iceland. J. Cetacean Res. Manage. 5:125-129. Jeffrey, C. F. G. 2000. Annual, coastal and seasonal variation in Grenadian demersal fisheries (1986 -1993) and implications for management. Bull. Mar. Sci. 66:305-319. Johns, W.E., Townsend, T.L., Fratantoni, D.M., and Wilson, W.D. 2002. On the Atlantic inflow to the Caribbean Sea. Deep-Sea Res. (1 Oceanogr. Res. Pap.) 49:211-243. Kaschner, K. 2004. Modelling and mapping of resource overlap between marine mammals and fisheries on a global scale. MMRU, Fisheries Centre, Department of Zoology. University of British Columbia, Vancouver, Canada. PhD Thesis. pp 184 Kaschner, K., Watson, R., Trites, A. W., and Pauly, D. 2006. Mapping worldwide distributions of marine mammals using a Relative Environmental Suitability (RES) model. Mar. Ecol. Prog. Ser..316:285-310. Katona, S. K. and Beard, J. A. 1990. Population size. migrations and feeding aggregations of the humpback whale (Megaptera novaeangliae) in the western North Atlantic Ocean. Rep. Int. Whaling Comm. 12:295-305. Kawakami, T. 1980. A review of sperm whale food. Sci. Rep. Whales Res. Inst. (Tokyo) 32:199-218. Klinowska, M. 1991. Dolphins, porpoises, and whales of the world: the IUCN red data book. IUCN - The World Conservation Union, Gland, Switzerland. Lavigne, D.M. 1996. Ecological interactions between marine mammals, commercial fisheries and their prey: unravelling the tangled web. In W. A. Montevecchi, (Ed). Studies of high latitude seabirds. 4. Trophic relationships and energetics of endotherms in cold ocean systems. Canadian Wildlife Service Occasional Paper Ottawa. Lenfest Ocean Program. Available at World Wide Web http://www.lenfestocean.org/, accessed on September 30, 2008. Lockyer, C. 1981. Growth and energy budgets of large baleen whales from the southern hemisphere. Mammals in the Seas: General Papers and Large Cetaceans. FAO Fisheries Series 5:379-487. Rome. Lydersen, C., Weslawski, J.M., and Oritsland, N.A. 1991. Stomach content analysis of minke whales Balaenoptera acutorostrata from the Lofoten and Vesteralen areas, Norway. Holarctic Ecol. 14:219-222. Mackinson, S., Blanchard, J. L., Pinnegar, J. K., and Scott, R. 2003. Consequences of alternative functional response formations in models exploring whale-fishery interactions. Mar. Mammal Sci. 19:661-681. Mackinson, S., Okey, T.A., Vasconcellos, M., Vidal-Hernandez, L., and Mahmodi, B. 2005. An ecosystem model of west Florida shelf for use in fisheries management and ecological Research. A report prepared for the Florida Marine Research Institute, St Petersburg, Florida. Mahon, R. 1989. Developing a management strategy for the flyingfish fishery of the Eastern Caribbean. Proc. Gulf Caribb. Fish. Instit. 39:389-402. Mahon, R. 1990. Fishery management options for Lesser Antilles countries. FAO Fisheries Technical Paper. FAO, Rome. pp 126 Mahon, R. 1993. Lesser Antilles. Marine fishery resources of the Antilles. FAO Fish. Tech. Pap. 326:5-98. Mahon, R. 1996. Fisheries and research for tunas and tuna-like species in the Western Central Atlantic. FAO Fish. Tech. Pap. 357:62. Mahon, R. 1999. Dolphinfish fisheries in the Caribbean region. Sci. Mar. 63:411-420. Mahon, R. and McConney, P.A. 2004. Management of large pelagic fisheries in CARICOM countries. FAO, Rome. Mahon, R. and Oxenford, H.A. 1999. Precautionary assessment and management of dolphinfish in the Caribbean. Sci. Mar. 63:429- 438. Mahon, R., Rennie, J., Ryan, R., and Singh-Renton, S. 1994a. Billfish catch and effort data from Barbados, Grenanda, St. Lucia and St. Vincent and the Grenadines. Col. Vol. Sci. Pap. ICCAT 41:431-441. Mahon, R., Singh-Renton, S., Jennings-Clarke, S., Rennie, J., Ryan, R., and Willoughby, S. 1994b. Yellowfin tuna catch and effort data from Barbados, Grenada, St. Lucia and St. Vincent and the Grenadines. Col.Vol.Sci. Pap. ICCAT 42:199-203. Marcano, J.S., Lárez, A., and Gutiérrez, X. 2002. Pesquería de tunidos por pequeños palangreros en el mar caribe y el océano altlántico durante el período 1986-2000. Bol. Inst. Oceanogr. Venezuela, Univ. Oriente 41:73-82. Marcano, J.S., Larez, A., Gutierrez, X., Alio, J. J., Salazar, H., and Marquez, M. 2004. Incidental catch of billfish and other species by Venezuelan longline vessels in the Caribbean Sea and western Atlantic Ocean: period 1986-2000. Cienc. Mar. 30:201-217. Marcano, L.A., Alió, J.J., Arocha, F., and Gutiérrez, X. 2001. Tendencia actual de la persquería artesanal de peces de pico en La Costa Central de Venezuela. período 1988-1999 SCRS/20002074. Col. Vol. Sci. Pap. ICCAT 53:281-290. Martin, C.S., Jeffers, J., and Godley, B.J. 2005. The status of marine turtles in Montserrat (Eastern Caribbean). Anim. Biodivers. Conserv. 28:159-168. Matos-Caraballo, D., Posada, J.M., and Luckhurst, B. 2006. Fishery-depedent evaluation of a spawning aggregation of tiger grouper (Mycteroperca tigris) at Vieques Island, Puerto Rico. Bull. Mar. Sci. 79:1-16. Melvin, G., Fanning, P., O'Donnell, C., Dahl, M., Edwards, L., Gardner, R., Simon, H., and Theophille, D. 2007. Scientific basis for ecosystem-based management in the Lesser Antilles including interactions with marine mammals and other top predators: acoustic biomass estimates of pelagic forage species in the offshore waters of the Lesser Antilles. FAO, Barbados. Mendoza, J.J. and Larez, A. 2004. A biomass dynamics assessment of the southeastern Caribbean snapper-grouper fishery. Fish. Res. 66:129-144. Food web models and data for the Caribbean model, J.L. Melgo et al.  114 Mitchell, E. D. 1973. Draft report on humpback whales taken under special scientific permit by eastern Canadian land stations. Rep. Int. Whal. Commn. 32:161-169. Mohammed, E. 2003a. Part II: ecosystem models - A generic marine ecosystem model for the southeastern Caribbean in the late 1990s: application to Grenada and the Grenadines. Fish. Centre Res. Rep. 11(6):191-225. The Fisheries Centre, University of British Columbia, Vancouver, B.C., Canada. Mohammed, E. 2003b. Reconstructing fisheries catches and fishing effort for the southeastern Caribbean (1940-2001): general methodology. pp 11-20. In D. Zeller, Booth, S., Mohammed, E., and Pauly, D., (Eds). From Mexico to Brazil: Central Atlantic fisheries catch trends and ecosystem models. Fish. Centre Res. Rep. 9. The Fisheries Centre, University of British Columbia, Vancouver, Canada. Mohammed, E., Fanning, P., Parker, C., Theophille, D., Martin, L., Punnett, S., Wilkins, R., Rambally, J., Philip, P., Isaac, C., Philmore, J., and Barrett, A. 2007a. Scientific basis for ecosystem-based management in the Lesser Antilles including interactions with marine mammals and other top predators: estimated catch, price and value for national fleet sectors from pelagic fisheries in the Lesser Antilles, Rome, Italy. Mohammed, E., Vasconcellos, M., Mackinson, S., Fanning, P., Heileman, S., and Carocci, F. 2007b. Scientific basis for ecosystem- based management in the Lesser Antilles including interactions with marine mammals and mther top predators: A trophic model of the Lesser Antilles Pelagic Ecosystem. FAO, Barbados. Morissette, L. 2005. Addressing uncertainty in marine ecosystems modelling. pp127-142. In E., Levner, I., Linkov and Proth, J.M. (Eds). NATO Science Series IV: Earth and Environmental SciencesNATO Science Series IV: Earth and Environmental Sciences. Springer Netherlands. Morissette, L., Hammill, M.O., and Savenkoff, C. 2006. The trophic role of marine mammals in the Northern Gulf of St. Lawrence. Mar. Mamm. Sci. 22:74-103. Morissette, L. 2007. Complexity, cost and quality of ecosystem models and their impact on resilience: a comparative analysis, with emphasis on marine mammals and the Gulf of St. Lawrence. Zoology. University of British Columbia, Vancouver, BC, Canada. PhD thesis. Morissette, L., Kaschner, K., and Gerber, L. submitted. Do whales deplete fisheries in Northwest African waters? Ecol. Appl. Morissette, L., Savenkoff, C., Castonguay, M., Swain, D., Chabot, D., Hammill, M.O., and Hanson, J.M. in press. Contrasting changes between the northern and southern Gulf of St. Lawrence ecosystems associated with the collapse of groundfish stocks. Deep Sea Res. II. Mullin, K.D. and Fulling, G.L. 2003. Abundance of cetaceans in the southern U.S. Atlantic ocean during summer 1998. Fishery Bulletin 101: 603-613. Mullin, K.D. and Fulling, G.L. 2004. Abundance of cetaceans in the oceanic northern Gulf of Mexico, 1996-2001. Marine Mammal Science 20: 787–807. Munro, J.L. 1983. Caribbean coral reef fishery resources. ICLARM Stud. Rev. 7:276 Murray, P.A. 1985. Growth and mortality in the dolphin-fish, Coryphaena hippurus, caught off Saint Lucia. W.I. FAO Fish. Rep. 327:147-151. Myers, R.A. and Worm, B. 2003. Rapid world-wide depletion of predatory fish communities. Nature 423:280-283. Nemeth, R. 2005. Population characteristics of a recovering US Virgin Islands red hind spawning aggregation following protection. Mar. Ecol. Prog. Ser. 286:81-97. Nørdoy, E.S. and Blix, A.S. 1992. Diet of minke whales in the northeastern Atlantic. Rep. Int. Whaling Comm.42:393-398. Northridge, S.P. 1984. World review of interactions between marine mammals and fisheries. FAO - Food and Agricultural Organisation of the United Nations, Rome. pp 190 Okey, T.A. 1999. Constructing and balancing the PWS model. pp 68-72. In T.A. Okey, and Pauly, D., (Eds). Trophic mass-balance model of Alaska’s Prince William Sound ecosystem, for the post-spill period 1994-1996. Fish. Centre Res. Rep. 7(4). The Fisheries Centre, University of British Columbia, Vancouver, B.C., Canada. Okey, T.A., Banks, S., Born, A.R., Bustamante, R.H., Calvopina, M., Edgar, G.J., Espinoza, E., Farina, J.M., Garske, L.E., Reck, G.K., Salazar, S., Shepherd, S., Toral-Granda, V., and Wallem, P. 2004. A trophic model of a Galapagos subtidal rocky reef for evaluating fisheries and conservation strategies. Ecol. Model. 172:383-401. Olsen, E. and Holst, J.C. 2001. A note on common minke whale (Balaenoptera acutorostrata) diets in the Norwegian Sea and the North Sea. J. Cetacean Res. Manage. 3:179-183. Olson, R. and Watters, G. 2003. A model of the pelagic ecosystem in the eastern tropical Pacific Ocean Bull. I-ATCC 22 (3):135-152. Opitz, S. 1996. Trophic interactions in Caribbean coral reefs. International Center for Living Aquatic Resources Management Technical Report, Makati City, Philippines. PhD thesis pp xiv + 341pp. Oxenford, H.A. 1985. Biology of the Dolphin Coryphaena hippurus and its implications for the Barbadian Fishery. The University of the West Indies, Cave Hill Campus, Barbados. PhD thesis. pp 366. Oxenford, H.A. 1999. Biology of the dolphinfish (Coryphaena hippurus) in the western central Atlantic: a review. Sci. Mar. 63:277- 301. Oxenford, H.A. and Hunte, W. 1986. A preliminary investigation of the stock structure of the dolphin, Coryphaena hippurus, in the Western Central Atlantic. Fish. Bull. 84:451-460. Oxenford, H.A. and Hunte, W. 1999. Feeding habits of the dolphinfish (Coryphaena hippurus) in the eastern Caribbean. Sci. Mar. 63:303-315. Oxenford, H. A., Mahon, R., and Hunt, H. 1995. Distribution and relative abundance of flyingfish (Exocoetidae) in the eastern Caribbean. I. Adults. Mar Ecol Prog Ser. 117:11-23. Oxenford, H. A., Murray, P. A., and Luckhurst, B. 2003. The biology of wahoo (Acanthocybium solandri) in the Western Central Atlantic. Proc. Gulf Caribb. Fish. Inst. 15:33-49. Paine, R.T., Tegner, M.J., and Johnson, E.A. 1998. Compounded perturbations yield ecological surprises. Ecosystems 1:535-545. Parker, C., Singh-Renton, S., Hackett, A., and Lauckner, F.B. 2001. Assessment of dolphinfish (Coryphaena hippurus) fishery using Eastern Caribbean data. CARICOM Fishery Report 9:41-53. Parker, C. 2002. Preliminary analysis of the flyingfish fishery of Barbados and interim recommendations. pp 156. In FAO/Western Central Atlantic Fishery Commission. Report of the Second Meeting of the WECAFC Ad Hoc Flyingfish Working Group of the Eastern Caribbean. Bridgetown, Barbados, 8-12 January 2001. FAO Fisheries Report, Rome, FAO. Pascoe, P. L., Mickiewicz, M. C., and Castello, H.P. 1990. Cephalopod remains from the stomach of a sperm whale stranded off Patagonia. Mar. Biol. 104:1-4. Pauly, D. 1980. On the interrelationships between natural mortality, growth parameters, and mean environmental temperature in 175 fish stocks. J. Cons.39:175-193. Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  115 Pauly, D., Soriano-Bartz, M.L., and Palomares, M.L.D. 1993. Improved construction, parametrization and interpretation of steady- state ecosystem models. pp 1-13. In V. Christensen, and Pauly, D., (Eds). Trophic models of aquatic ecosystems. ICLARM Conference Proceedings. Pauly, D., Christensen, V., Dalsgaard, J., Froese, R., and Torres, F. J. 1998a. Fishing down 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 J. Mar. Sci. 55:467-481. Pauly, D., Christensen, V., Guénette, S., Pitcher, T.J., Sumaila, U. R., Walters, C. J., Watson, R., and Zeller, D. 2002. Towards sustainability in world fisheries. Nature 418:689-695. Pauly, D., Christensen, V., and Haggan, N. 1996. Mass-balance model of Alaska Gyre. Fish. Centre Res. Rep. 4(1). The Fisheries Centre, University of British Columbia, Vancouver, B.C., Canada. Perrin, W. F., Würsig, B., and Thewissen, J.G.M. 2002. Encyclopedia of marine mammals. Academic Press, San Diego, CA. pp 1414 Perry, S.L., DeMaster, D.P., and Silber, G.K. 1999. The great whales: history and status of six species listed as endangered under the U.S. Endangered Species Act of 1973. Mar. Fish. Rev. 61:1-74. Pimm, S. L. 2002. Food Webs. University of Chicago Press. Plagányi, É.E. and Butterworth, D.S. 2002. Competition with fisheries. pp 268-273. In W.F. Perrin, Würsig, B., and Thewissen, J.G.M., (Eds). Encyclopedia of Marine Mammals. Academic Press, San Diego, CA. Plagányi, É.E. and Butterworth, D.S. 2004. A critical look at the potential of ECOPATH with ECOSIM to assist in practical fisheries management. Afr. J. Mar. Sci. 26:261-287. Plagányi, É.E., Rademeyer, R.A., Butterworth, D.S., Cunningham, C.L., and Johnston, S.J. 2007. Making management procedures operational - innovations implemented in South Africa. ICES J. Mar. Sci. 64:626-632. Potts, A.C., Thomas, A.D., and Nichols, E. 2002. An economic and social assessment of the flyingfish (pelagic) fishery of Trinidad and Tobago. FAO/Western Central Atlantic Fishery Commission. Report of the second meeting of the WECAFC Ad Hoc Flyingfish Working Group of the Eastern Caribbean. FAO Fisheries Report Bridgetown, Barbados. Prager, M.H. 2000. Exploratory assessment of dolphinfish, Coryphaena hippurus, based on US landings from the Atlantic Ocean and Gulf of Mexico. Laboratory Document, NMFS, NOAA, Beaufort, NC. pp 18. Priddel, D., Hutton, I., Olsen, S., and Wheeler, R. 2005. Breeding biology of masked boobies (Sula dactylatra tasmani) on Lord Howe Island, Australia. Emu. 105:105-113. Punt, A.E. and Butterworth, D.S. 1995. The effects of future consumption by Cape fur seal on catches and catch rates of the Cape hakes. 4. Modelling the biological interaction between Cape fur seals Arctocephalus pusillus pusillus and the Cape hake Merluccius capensis and Merluccius paradoxus. S. Afr. J. Mar. Sci. 16:255-285. Quimby, S. 2000. Whaling. Caribbean Currents 8(2):1-8. Rambally, J. 1999. Whaling in St Lucia: a dying tradition? Department of Fisheries, Ministry of Agriculture, Forestry, Fisheries and the Environment, St Lucia. Rajendra, M.D.K., Khan, A.A., Knight, D., O'Reilly, O., Chang Yen, I., Wagh, A.B. and Desai, B.N. 1991. Some aspects of nutrient chemistry of the Caribbean Sea. Caribb. Mar. Stud. 2:81-86. Read, A.J., Halpin, P.N., Crowder, L.B., Best, B.D., and Fujioka, E. 2007. OBIS-SEAMAP: mapping marine mammals, birds and turtles. Recksiek, C.W., Appeldoorn, R.S., and Turingan, R.G. 1991. Studies of fish traps as stock assessment devices on a shallow reef in south-western Puerto Rico. Fish. Res. 10:177-197. Reeves, R. R. 2002. The origins and character of ‘aboriginal subsistence’ whaling: a global review. Mammal Rev. 32:71–106. Reeves, R.R., Swartz, S.L., Wetmore, S.E., and Clapham, P.J. 2001. Historical occurrence and distribution of humpback whales in the eastern and southern Caribbean Sea, based on data from American whaling logbooks. J. Cetacean Res. Manage. 3:117-129. Restrepo, V., Prince, E.D., Scott, G.P., and Uozumi, Y. 2003. ICCAT stock assessments of Atlantic billfish. Mar. Freshwater Res. 54:361-367. Richards, W.J. and Bohnsack, J.A. 1990. The Caribbean Sea: a large marine ecosystem in crisis. pp 44-53. In Sherman, K., Alexander, L.M., and Gold, B.D., (Eds). Large Marine Ecosystems: Stress, Mitigation, and Sustainability. American Association for the Advancement of Science, Washington, D.C. Romero, A., Agudo, A. I., Green, S. M., and Notarbartolo-di-Sciara, G. 2001. Cetaceans of Venezuela: Their Distribution and Conservation Status. NMFS. pp 60 Santos, M.B., Martin, V., Arbelo, M., Fernández, A., and Pierce, G.J. 2007. Insights into the diet of beaked whales from the atypical mass stranding in the Canary Islands in September 2002. J. Mar. Biol. Assoc. U. K. 87:243-251. Santos, M. B., Pierce, G. J., Herman, J., López, A., Guerra, A., Mente, E., and Clark, M. R. 2001. Feeding ecology of Cuvier's beaked whale (Ziphius cavirostris): a review with new information on the diet of this species. J. Mar. Biol. Assoc. U. K.  81:687- 694. Santos, M.B., Pierce, G.J., Lopez, A., Reid, R.J., Ridoux, V., and Mente, E. 2006. Pygmy sperm whales Kogia breviceps in the northeast Atlantic: New information on stomach contents and strandings. Mar. Mamm. Sci. 22:600-616. Saul, S., Diaz, G., and Rosario, A. 2005. Preliminary analysis and standardized catch per unit effort indices for yellowtail snapper fishery independent data in Puerto Rico in SEDAR, editor. SEDAR 8 Stock Assessment Report. Caribbean Spiny Lobsters and Yellowtail Snapper. Southeast Fisheries Science Center, NOAA, Charleston, SC. Savenkoff, C., Swain, D.P., Hanson, J.M., Castonguay, M., Hammill, M.O., Bourdages, H., Morissette, L., and Chabot, D. 2007. Effects of fishing and predation in a heavily exploited ecosystem: comparing pre- and post-groundfish collapse periods in the southern Gulf of St. Lawrence (Canada). Ecol. Model. 204:115-128. Schreiber, E.A. and Lee, D.S. 2000. West Indian seabirds: a disappearing natural resources. In E.A. Schreiber, and Lee, D. S., (Eds). Status and Conservation of West Indian Seabirds. Society of Caribbean Ornithology, Special Publication, Ruston, Louisiana. Sears, R. and Harriet, H. 2002. Blue Whale - Balaenoptera musculus. pp 112-122. In W.F. Perrin, Würsig, B., and Thewissen, H.G.M., (Eds). Encyclopedia of Marine Mammals. Academic Press, San Diego, CA. Sea Around Us Project (SAUP). Available at World Wide Web http://www.seaaroundus.org/, accessed on August 15, 2008. SEDAR. 2005a. SEDAR 8 Stock assessment report I: Caribbean yellowtail snapper, Charleston, SC. SEDAR. 2005b. SEDAR 8 Stock assessment report II: Caribbean spiny lobster, Charleston, SC. SEDAR. 2007. Caribbean queen conch. SEDAR 14 stock assessment report 3. Southeast Fisheries Science Center and the South Atlantic Fishery Management Council, North Charleston, SC. Food web models and data for the Caribbean model, J.L. Melgo et al.  116 Sidi, M.T. and Guénette, S. 2004. Modèle trophique de la ZEE mauritanienne: comparaison de deux périodes (1987 et 1998). pp 12- 38. In M.L.D. Palomares, and Pauly, D., (Eds). West African marine ecosystems: models and fisheries impacts. Fish. Centre Res. Rep. 12(7). The Fisheries Centre, University of British Columbia, Vancouver, B.C., Canada. Sivertsen, S.P., Pedersen, T., Lindstrom, U.L.F., and Haug, T. 2006. Prey partitioning between cod (Gadus morhua) and minke whale (Balaenoptera acutorostrata) in the Barents Sea. Mar. Biol. Res. 2:89-99. Smith, S. C. and Whitehead, H. 2000. The diet of Galapagos sperm whales Physeter macrocephalus as indicated by fecal sample analysis. Mar. Mamm. Sci, 16:315-325. Smout, S. and Lindstrom, U. 2007. Multispecies functional response of the minke whale Balaenoptera acutorostrata based on small- scale foraging studies. Mar. Ecol. Prog. Ser. 341:277-291. Swartz, S.L., Cole, T., McDonald, M.A., Hildebrand, J.A., Oleson, E.M., Martinez, A., Clapham, P. J., Barlow, J., and Jones, M. L. 2003. Acoustic and visual survey of humpback whale (Megaptera novaeangliae) distribution in the eastern and southeastern Caribbean Sea. Caribb. J. Sci. 39:195-208. Tamura, T. 2003. Regional assessment of prey consumption and competition by marine cetaceans in the world. Pp 143-170. In M. Sinclair, and Valdimarsson, G., (Eds). Responsible Fisheries in Marine Ecosystems. Food and Agricultural Organisation of the United Nations and CABI Publishing, Wallingford, UK. Tershy, B. R. 1992. Body size, diet, habitat use, and social behavior of Balaenoptera whales in the Gulf of California. J. Mammal. 73:477-486. Trites, A. and Heise, K. 1996. Marine Mammals. pp 25-30. In D. Pauly, Christensen, V., and Haggan, N., editors. Mass-balance models of North-eastern Pacific ecosystems. Fish. Centre Res. Rep. 4(1). The Fisheries Centre, University of British Columbia, Vancouver, B.C., Canada. Trites, A.W., Christensen, V., and Pauly, D. 1997. Competition between fisheries and marine mammals for prey and primary production in the Pacific Ocean. J. Northwest. Atl. Fish. Sci. 22:173-187. Trites, A.W. and Pauly, D. 1998. Estimating mean body masses of marine mammals from maximum body lengths. Can. J. Zool. 76:886-896. UNEP. 2002. Activities of the Caribbean Environment Programme (CEP) of UNEP on the conservation of marine mammals in the wider Caribbean, Monaco. Vasconcellos, M. and Watson, R. 2004. Mass balance of Atlantic oceanic systems. pp 171-214. In M. L. D. Palomares, and Pauly, D., (Eds). West African marine ecoystems: models and fisheries impacts. Fish. Centre Res. Rep. 12(7). The Fisheries Centre, University of British Columbia, Vancouver, B.C., Canada. Vidal, L. and Basurto, M. 2003. A preliminary trophic model of Bahía de la Ascensión, Quintana Roo, Mexico [modelo trófico preliminar de Bahía Ascención, Quintana Roo, México]. Fish. Centre Res. Rep. 11:255-264. Wade, P.R. and Gerrodette, T. 1993. Estimates of cetacean abundance and distribution in the Eastern Tropical Pacific. Reports of the International Whaling Commission 43: 477-493. Walters, C. and Kitchell, J.F. 2004. Fisheries ecology and management. Princeton University Press, Princeton, New Jersey. Watson, R., Kitchingman, A., Gelchu, A., and Pauly, D. 2004. Mapping global fisheries: sharpening our focus. Fish Fish. 5:168-177. Windsland, K., Lindstrom, U., Nilssen, K.T., and Haug, T. 2007. Relative abundance and size composition of prey in the common minke whale diet in selected areas of the northeastern Atlantic during 2000-04. J. Cetacean Res. Manage. 9:167-178. Witteveen, B.H., Foy, R.J., and Wynne, K.M. 2006. The effect of predation (current and historical) by humpbak whales (Megaptera novaeangliae) on fish abundance near Kodiak Island, Alaska. Fish. Bull. 104:10-20. Wolff, M., Koch, V., Chavarria, J.B, and Vargas, J.A. 1998. A trophic flow model of the Golfo de Nicoya, Costa Rica. Rev. Biol. Mar. Tropical 46:63-79. Zeller, D., Booth, S., Mohammed, E., and Pauly, D. 2003. From Mexico to Brazil: Central Atlantic Fisheries Catch Trends and Ecosystem Models. Fish. Centre Res. Rep. 11(6). The Fisheries Centre, University of British Columbia, Vancouver, B.C., Canada. Zeller, D. and Freire, K. 2001. A north-east Atlantic marine ecosystem model for the Faroe Islands (ICES Area VB): input data. Fish. Centre Res. Rep. 9:207-212. The Fisheries Centre, University of British Columbia, Vancouver, B.C., Canada.   Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  117 APPENDICES Appendix 1. Results of the sensitivity analysis for the Northwest African model. Input and estimated parameter numbers correspond to Ecopath groups.     Change in input parameters Input parameter Estimated parameter -50% -40% -30% -20% -10% 0% 10% 20% 30% 40% 50% 5 B 8 EE -0.500 -0.400 -0.300 -0.200 -0.100  0  0.100 0.200 0.300 0.400 0.500 5 Q/B 8 EE -0.500 -0.400 -0.300 -0.200 -0.100  0  0.100 0.200 0.300 0.400 0.500 8 B 8 EE 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 8 P/B 8 EE 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 9 B 17 B -0.154 -0.123 -0.092 -0.062 -0.031  0  0.031 0.062 0.092 0.123 0.154 9 B 18 B -0.188 -0.150 -0.113 -0.075 -0.038  0  0.038 0.075 0.113 0.150 0.188 9 B 23 EE -0.146 -0.117 -0.087 -0.058 -0.029  0  0.029 0.058 0.087 0.117 0.146 9 Q/B 17 B -0.154 -0.123 -0.092 -0.062 -0.031  0  0.031 0.062 0.092 0.123 0.154 9 Q/B 18 B -0.188 -0.150 -0.113 -0.075 -0.038  0  0.038 0.075 0.113 0.150 0.188 9 Q/B 23 EE -0.146 -0.117 -0.087 -0.058 -0.029  0  0.029 0.058 0.087 0.117 0.146 10 P/B 10 B 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 10 P/B 17 B 0.062 0.041 0.026 0.015 0.007  0  -0.006 -0.010 -0.014 -0.018 -0.021 10 P/B 18 B 0.140 0.093 0.060 0.035 0.016  0  -0.013 -0.023 -0.032 -0.040 -0.047 10 P/B 19 B 0.242 0.161 0.104 0.060 0.027  0  -0.022 -0.040 -0.056 -0.069 -0.081 10 P/B 23 EE 0.091 0.061 0.039 0.023 0.010  0  -0.008 -0.015 -0.021 -0.026 -0.030 10 Q/B 18 B -0.070 -0.056 -0.042 -0.028 -0.014  0  0.014 0.028 0.042 0.056 0.070 10 Q/B 19 B -0.121 -0.097 -0.073 -0.048 -0.024  0  0.024 0.048 0.073 0.097 0.121 10 Q/B 23 EE -0.046 -0.037 -0.027 -0.018 -0.009  0  0.009 0.018 0.027 0.037 0.046 10 EE 10 B 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 10 EE 17 B 0.062 0.041 0.026 0.015 0.007  0  -0.006 -0.010 -0.014 -0.018 -0.021 10 EE 18 B 0.140 0.093 0.060 0.035 0.016  0  -0.013 -0.023 -0.032 -0.040 -0.047 10 EE 19 B 0.242 0.161 0.104 0.060 0.027  0  -0.022 -0.040 -0.056 -0.069 -0.081 10 EE 23 EE 0.091 0.061 0.039 0.023 0.010  0  -0.008 -0.015 -0.021 -0.026 -0.030 11 B 10 B -0.259 -0.208 -0.156 -0.104 -0.052  0  0.052 0.104 0.156 0.208 0.259 11 B 12 B -0.496 -0.397 -0.297 -0.198 -0.099  0  0.099 0.198 0.297 0.397 0.496 11 B 16 B -0.056 -0.045 -0.034 -0.023 -0.011  0  0.011 0.022 0.034 0.045 0.056 11 B 17 B -0.037 -0.030 -0.022 -0.015 -0.007  0  0.007 0.015 0.022 0.030 0.037 11 B 18 B -0.044 -0.035 -0.026 -0.017 -0.009  0  0.009 0.017 0.026 0.035 0.044 11 B 19 B -0.185 -0.148 -0.111 -0.074 -0.037  0  0.037 0.074 0.111 0.148 0.185 11 B 20 EE -0.496 -0.397 -0.297 -0.198 -0.099  0  0.099 0.198 0.297 0.397 0.496 11 B 23 EE -0.042 -0.034 -0.025 -0.017 -0.008  0  0.008 0.017 0.025 0.034 0.042 11 P/B 11 EE 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 11 Q/B 10 B -0.259 -0.208 -0.156 -0.104 -0.052  0  0.052 0.104 0.156 0.208 0.259 11 Q/B 11 EE -0.500 -0.400 -0.300 -0.200 -0.100  0  0.100 0.200 0.300 0.400 0.500 11 Q/B 12 B -0.496 -0.397 -0.297 -0.198 -0.099  0  0.099 0.198 0.297 0.397 0.496 11 Q/B 16 B -0.056 -0.045 -0.034 -0.023 -0.011  0  0.011 0.022 0.034 0.045 0.056 11 Q/B 17 B -0.037 -0.030 -0.022 -0.015 -0.007  0  0.007 0.015 0.022 0.030 0.037 11 Q/B 18 B -0.044 -0.035 -0.026 -0.017 -0.009  0  0.009 0.017 0.026 0.035 0.044 11 Q/B 19 B -0.185 -0.148 -0.111 -0.074 -0.037  0  0.037 0.074 0.111 0.148 0.185 11 Q/B 20 EE -0.496 -0.397 -0.297 -0.198 -0.099  0  0.099 0.198 0.297 0.397 0.496 11 Q/B 23 EE -0.042 -0.034 -0.025 -0.017 -0.008  0  0.008 0.017 0.025 0.034 0.042 12 P/B 12 B 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 12 P/B 16 B 0.060 0.040 0.026 0.015 0.007  0  -0.005 -0.010 -0.014 -0.017 -0.020 12 P/B 20 EE 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 12 Q/B 20 EE -0.500 -0.400 -0.300 -0.200 -0.100  0  0.100 0.200 0.300 0.400 0.500 12 EE 12 B 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 12 EE 16 B 0.060 0.040 0.026 0.015 0.007  0  -0.005 -0.010 -0.014 -0.017 -0.020 12 EE 20 EE 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 13 B 13 EE 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 13 P/B 13 EE 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 14 B 14 EE 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 14 P/B 14 EE 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 15 B 15 EE 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 15 B 17 B -0.147 -0.118 -0.088 -0.059 -0.029  0  0.029 0.059 0.088 0.118 0.147 15 B 18 B -0.123 -0.099 -0.074 -0.049 -0.025  0  0.025 0.049 0.074 0.099 0.123 15 B 19 B -0.164 -0.131 -0.098 -0.066 -0.033  0  0.033 0.066 0.098 0.131 0.164 15 B 23 EE -0.147 -0.117 -0.088 -0.059 -0.029  0  0.029 0.059 0.088 0.117 0.147 15 P/B 15 EE 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 15 Q/B 17 B -0.147 -0.118 -0.088 -0.059 -0.029  0  0.029 0.059 0.088 0.118 0.147 15 Q/B 18 B -0.123 -0.099 -0.074 -0.049 -0.025  0  0.025 0.049 0.074 0.099 0.123 15 Q/B 19 B -0.164 -0.131 -0.098 -0.066 -0.033  0  0.033 0.066 0.098 0.131 0.164  118 15 Q/B 23 EE -0.147 -0.117 -0.088 -0.059 -0.029  0  0.029 0.059 0.088 0.117 0.147 16 P/B 16 B 1.005 0.669 0.430 0.251 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.334 16 EE 16 B 1.005 0.669 0.430 0.251 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.334 17 P/B 17 B 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 17 P/B 23 EE 0.672 0.448 0.288 0.168 0.075  0  -0.061 -0.112 -0.155 -0.192 -0.224 17 Q/B 23 EE -0.336 -0.269 -0.202 -0.134 -0.067  0  0.067 0.134 0.202 0.269 0.336 17 EE 17 B 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 17 EE 23 EE 0.672 0.448 0.288 0.168 0.075  0  -0.061 -0.112 -0.155 -0.192 -0.224 18 P/B 18 B 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 18 P/B 23 EE 0.224 0.150 0.096 0.056 0.025  0  -0.020 -0.037 -0.052 -0.064 -0.075 18 Q/B 23 EE -0.112 -0.090 -0.067 -0.045 -0.022  0  0.022 0.045 0.067 0.090 0.112 18 EE 18 B 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 18 EE 23 EE 0.224 0.150 0.096 0.056 0.025  0  -0.020 -0.037 -0.052 -0.064 -0.075 19 P/B 16 B 0.123 0.082 0.052 0.031 0.014  0  -0.011 -0.020 -0.028 -0.035 -0.041 19 P/B 18 B 0.058 0.039 0.025 0.014 0.006  0  -0.005 -0.010 -0.013 -0.016 -0.019 19 P/B 19 B 1.005 0.669 0.430 0.251 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.334 19 P/B 23 EE 0.060 0.040 0.026 0.015 0.007  0  -0.005 -0.010 -0.014 -0.017 -0.020 19 Q/B 16 B -0.061 -0.049 -0.037 -0.024 -0.012  0  0.012 0.024 0.037 0.049 0.061 19 EE 16 B 0.123 0.082 0.052 0.031 0.014  0  -0.011 -0.020 -0.028 -0.035 -0.041 19 EE 18 B 0.058 0.039 0.025 0.014 0.006  0  -0.005 -0.010 -0.013 -0.016 -0.019 19 EE 19 B 1.005 0.669 0.430 0.251 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.334 19 EE 23 EE 0.060 0.040 0.026 0.015 0.007  0  -0.005 -0.010 -0.014 -0.017 -0.020 20 B 20 EE 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 20 B 21 B -0.458 -0.367 -0.275 -0.183 -0.092  0  0.092 0.183 0.275 0.367 0.458 20 P/B 20 EE 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 20 Q/B 21 B -0.458 -0.367 -0.275 -0.183 -0.092  0  0.092 0.183 0.275 0.367 0.458 21 P/B 21 B 1.786 1.053 0.625 0.345 0.147  0  -0.114 -0.204 -0.278 -0.339 -0.391 21 Q/B 21 B -0.124 -0.101 -0.078 -0.053 -0.027  0  0.029 0.060 0.092 0.127 0.164 21 EE 21 B 1.786 1.053 0.625 0.345 0.147  0  -0.114 -0.204 -0.278 -0.339 -0.391 23 B 23 EE 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 23 B 24 Q/B -0.490 -0.392 -0.294 -0.196 -0.098  0  0.098 0.196 0.294 0.392 0.490 23 P/B 23 EE 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 23 Q/B 24 Q/B -0.490 -0.392 -0.294 -0.196 -0.098  0  0.098 0.196 0.294 0.392 0.490 24 B 24 Q/B 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333 24 P/B 24 Q/B 1.000 0.667 0.429 0.250 0.111  0  -0.091 -0.167 -0.231 -0.286 -0.333  Modelling the trophic role of marine mammals in tropical areas, L. Morissette et al.  119  Appendix 2. Results of the sensitivity analysis. Input and estimated parameter numbers correspond to Ecopath groups.     Change in input parameters Input parameter Estimated parameter -50% -40% -30% -20% -10% 0% 10% 20% 30% 40% 50% 1 B 14 EE -0.071 -0.057 -0.043 -0.029 -0.014 0 0.014 0.029 0.043 0.057 0.071 1 Q/B 14 EE -0.071 -0.057 -0.043 -0.029 -0.014 0 0.014 0.029 0.043 0.057 0.071 3 B 14 EE - 0.060 -0.048 -0.036 -0.024 -0.012 0 0.012 0.024 0.036 0.048 0.060 3 Q/B 14 EE - 0.060 -0.048 -0.036 -0.024 -0.012 0 0.012 0.024 0.036 0.048 0.060 4 B 14 EE -0.325 -0.260 -0.195 -0.130 -0.065 0 0.065 0.130 0.195 0.260 0.325 4 Q/B 14 EE -0.325 -0.260 -0.195 -0.130 -0.065 0 0.065 0.130 0.195 0.260 0.325 7 B 10 EE -0.500 -0.400 -0.300 -0.200 -0.100 0 0.100 0.200 0.300 0.400 0.500 7 Q/B 10 EE -0.500 -0.400 -0.300 -0.200 -0.100 0 0.100 0.200 0.300 0.400 0.500 10 B 10 EE 1.000 0.667 0.429 0.250 0.111 0 -0.091 -0.167 -0.231 -0.286 -0.333 10 B 13 EE -0.193 -0.154 -0.116 -0.077 -0.039 0 0.039 0.077 0.116 0.154 0.193 10 P/B 10 EE 1.000 0.667 0.429 0.250 0.111 0 -0.091 -0.167 -0.231 -0.286 -0.333 10 Q/B 13 EE -0.193 -0.154 -0.116 -0.077 -0.039 0 0.039 0.077 0.116 0.154 0.193 11 B 11 EE 1.000 0.667 0.429 0.250 0.111 0 -0.091 -0.167 -0.231 -0.286 -0.333 11 B 16 EE -0.153 -0.122 -0.092 -0.061 -0.031 0 0.031 0.061 0.092 0.122 0.153 11 P/B 11 EE 1.000 0.667 0.429 0.250 0.111 0 -0.091 -0.167 -0.231 -0.286 -0.333 11 Q/B 16 EE -0.153 -0.122 -0.092 -0.061 -0.031 0 0.031 0.061 0.092 0.122 0.153 12 B 12 EE 1.000 0.667 0.429 0.250 0.111 0 -0.091 -0.167 -0.231 -0.286 -0.333 12 P/B 12 EE 1.000 0.667 0.429 0.250 0.111 0 -0.091 -0.167 -0.231 -0.286 -0.333 13 B 13 EE 0.413 0.275 0.177 0.103 0.046 0 - 0.038 -0.069 -0.095 -0.118 -0.138 13 B 17 EE -0.039 -0.031 -0.023 -0.016 - 0.008 0 0.008 0.016 0.023 0.031 0.039 13 P/B 13 EE 1.000 0.667 0.429 0.250 0.111 0 -0.091 -0.167 -0.231 -0.286 -0.333 13 Q/B 13 EE -0.294 -0.235 -0.176 -0.117 -0.059 0 0.059 0.117 0.176 0.235 0.294 13 Q/B 17 EE -0.039 -0.031 -0.023 -0.016 - 0.008 0 0.008 0.016 0.023 0.031 0.039 14 B 14 EE 1.000 0.667 0.429 0.250 0.111 0 -0.091 -0.167 -0.231 -0.286 -0.333 14 B 23 EE -0.081 -0.064 -0.048 -0.032 -0.016 0 0.016 0.032 0.048 0.064 0.081 14 P/B 14 EE 1.000 0.667 0.429 0.250 0.111 0 -0.091 -0.167 -0.231 -0.286 -0.333 14 Q/B 23 EE -0.081 -0.064 -0.048 -0.032 -0.016 0 0.016 0.032 0.048 0.064 0.081 15 B 11 EE -0.500 -0.400 -0.300 -0.200 -0.100 0 0.100 0.200 0.300 0.400 0.500 15 B 12 EE -0.115 -0.092 -0.069 -0.046 -0.023 0 0.023 0.046 0.069 0.092 0.115 15 B 15 EE 0.233 0.155 0.100 0.058 0.026 0 -0.021 -0.039 -0.054 -0.066 -0.078 15 B 16 EE -0.099 -0.079 -0.059 -0.040 -0.020 0 0.020 0.040 0.059 0.079 0.099 15 P/B 15 EE 1.000 0.667 0.429 0.250 0.111 0 -0.091 -0.167 -0.231 -0.286 -0.333 15 Q/B 11 EE -0.500 -0.400 -0.300 -0.200 -0.100 0 0.100 0.200 0.300 0.400 0.500 15 Q/B 12 EE -0.115 -0.092 -0.069 -0.046 -0.023 0 0.023 0.046 0.069 0.092 0.115 15 Q/B 15 EE -0.384 -0.307 -0.230 -0.154 -0.077 0 0.077 0.154 0.230 0.307 0.384 15 Q/B 16 EE -0.099 -0.079 -0.059 -0.040 -0.020 0 0.020 0.040 0.059 0.079 0.099 16 B 16 EE 0.573 0.382 0.245 0.143 0.064 0 -0.052 -0.095 -0.132 -0.164 -0.191 16 B 19 EE -0.035 -0.028 -0.021 -0.014 -0.007 0 0.007 0.014 0.021 0.028 0.035 16 P/B 16 EE 1.000 0.667 0.429 0.250 0.111 0 -0.091 -0.167 -0.231 -0.286 -0.333 16 Q/B 16 EE -0.214 -0.171 -0.128 -0.086 -0.043 0 0.043 0.086 0.128 0.171 0.214 16 Q/B 19 EE -0.035 -0.028 -0.021 -0.014 -0.007 0 0.007 0.014 0.021 0.028 0.035 17 B 17 EE 0.793 0.528 0.340 0.198 0.088 0 -0.072 -0.132 -0.183 -0.226 -0.264 17 P/B 17 EE 1.000 0.667 0.429 0.250 0.111 0 -0.091 -0.167 -0.231 -0.286 -0.333 17 Q/B 17 EE -0.104 -0.083 -0.062 -0.042 -0.021 0 0.021 0.042 0.062 0.083 0.104 18 P/B 12 EE 2.488 1.230 0.667 0.349 0.143 0 -0.106 -0.187 -0.251 -0.303 -0.346 18 P/B 15 EE 0.684 0.338 0.183 0.096 0.039 0 -0.029 -0.051 -0.069 - 0.083 -0.095 18 P/B 16 EE 0.177 0.087 0.047 0.025 0.010 0 - 0.008 -0.013 -0.018 -0.022 -0.025 18 P/B 17 EE 0.606 0.300 0.163 0.085 0.035 0 -0.026 -0.046 -0.061 -0.074 - 0.084 18 P/B 19 EE 2.461 1.216 0.660 0.345 0.142 0 -0.105 -0.185 -0.248 - 0.300 -0.343 18 P/B 22 EE 0.777 0.384 0.208 0.109 0.045 0 -0.033 -0.058 -0.078 -0.095 -0.108 18 P/B 23 EE 1.352 0.669 0.363 0.189 0.078 0 -0.057 -0.102 -0.136 -0.165 -0.188 18 Q/B 12 EE -0.484 -0.404 -0.316 -0.221 -0.116 0 0.128 0.271 0.431 0.611 0.817 18 Q/B 15 EE -0.133 -0.111 -0.087 -0.061 -0.032 0 0.035 0.074 0.118 0.168 0.225 18 Q/B 16 EE -0.034 -0.029 -0.022 -0.016 - 0.008 0 0.009 0.019 0.031 0.043 0.058 18 Q/B 17 EE -0.118 -0.098 -0.077 -0.054 -0.028 0 0.031 0.066 0.105 0.149 0.199 18 Q/B 19 EE -0.479 -0.399 -0.313 -0.218 -0.114 0 0.127 0.268 0.426 0.605 0.808 18 Q/B 22 EE -0.151 -0.126 -0.099 -0.069 -0.036 0 0.040 0.085 0.135 0.191 0.255  120 18 Q/B 23 EE -0.263 -0.220 -0.172 -0.120 -0.063 0 0.070 0.147 0.234 0.332 0.444 19 B 19 EE 0.886 0.591 0.380 0.222 0.098 0 -0.081 -0.148 -0.204 -0.253 -0.295 19 P/B 19 EE 1.000 0.667 0.429 0.250 0.111 0 -0.091 -0.167 -0.231 -0.286 -0.333 19 Q/B 19 EE -0.057 -0.046 -0.034 -0.023 -0.011 0 0.011 0.023 0.034 0.046 0.057 20 B 17 EE -0.085 -0.068 -0.051 -0.034 -0.017 0 0.017 0.034 0.051 0.068 0.085 20 Q/B 17 EE -0.085 -0.068 -0.051 -0.034 -0.017 0 0.017 0.034 0.051 0.068 0.085 21 P/B 21 EE 1.000 0.667 0.429 0.250 0.111 0 -0.091 -0.167 -0.231 -0.286 -0.333 21 Q/B 21 EE -0.500 -0.400 -0.300 -0.200 -0.100 0 0.100 0.200 0.300 0.400 0.500 22 B 17 EE -0.126 -0.101 -0.075 -0.050 -0.025 0 0.025 0.050 0.075 0.101 0.126 22 B 19 EE -0.048 -0.039 -0.029 -0.019 -0.010 0 0.010 0.019 0.029 0.039 0.048 22 B 22 EE 0.884 0.589 0.379 0.221 0.098 0 - 0.080 -0.147 -0.204 -0.253 -0.295 22 B 23 EE -0.082 -0.066 -0.049 -0.033 -0.016 0 0.016 0.033 0.049 0.066 0.082 22 P/B 22 EE 1.000 0.667 0.429 0.250 0.111 0 -0.091 -0.167 -0.231 -0.286 -0.333 22 Q/B 17 EE -0.126 -0.101 -0.075 -0.050 -0.025 0 0.025 0.050 0.075 0.101 0.126 22 Q/B 19 EE -0.048 -0.039 -0.029 -0.019 -0.010 0 0.010 0.019 0.029 0.039 0.048 22 Q/B 22 EE -0.058 -0.047 -0.035 -0.023 -0.012 0 0.012 0.023 0.035 0.047 0.058 22 Q/B 23 EE -0.082 -0.066 -0.049 -0.033 -0.016 0 0.016 0.033 0.049 0.066 0.082 23 B 12 EE -0.292 -0.234 -0.175 -0.117 -0.059 0 0.059 0.117 0.175 0.234 0.292 23 B 15 EE - 0.080 -0.064 -0.048 -0.032 -0.016 0 0.016 0.032 0.048 0.064 0.080 23 B 17 EE -0.071 -0.057 -0.043 -0.029 -0.014 0 0.014 0.029 0.043 0.057 0.071 23 B 19 EE -0.289 -0.231 -0.174 -0.116 -0.058 0 0.058 0.116 0.174 0.231 0.289 23 B 22 EE -0.369 -0.295 -0.221 -0.147 -0.074 0 0.074 0.147 0.221 0.295 0.369 23 B 23 EE 0.552 0.368 0.237 0.138 0.061 0 -0.050 -0.092 -0.127 -0.158 -0.184 23 P/B 23 EE 1.000 0.667 0.429 0.250 0.111 0 -0.091 -0.167 -0.231 -0.286 -0.333 23 Q/B 12 EE -0.292 -0.234 -0.175 -0.117 -0.059 0 0.059 0.117 0.175 0.234 0.292 23 Q/B 15 EE - 0.080 -0.064 -0.048 -0.032 -0.016 0 0.016 0.032 0.048 0.064 0.080 23 Q/B 17 EE -0.071 -0.057 -0.043 -0.029 -0.014 0 0.014 0.029 0.043 0.057 0.071 23 Q/B 19 EE -0.289 -0.231 -0.174 -0.116 -0.058 0 0.058 0.116 0.174 0.231 0.289 23 Q/B 22 EE -0.369 -0.295 -0.221 -0.147 -0.074 0 0.074 0.147 0.221 0.295 0.369 23 Q/B 23 EE -0.224 -0.179 -0.134 - 0.090 -0.045 0 0.045 0.090 0.134 0.179 0.224    

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
United States 9 0
China 7 18
Greece 2 0
Brazil 2 0
Germany 1 3
Kuwait 1 0
City Views Downloads
Beijing 7 0
Unknown 6 7
Marlborough 2 0
Belo Horizonte 2 0
Kuwait City 1 0
Redmond 1 0
Athens 1 0
Mountain View 1 0
Ashburn 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}

Share

Share to:

Comment

Related Items