• Ecosystem service;
  • ecosystem-based management;
  • fisheries value;
  • forage fish;
  • supportive values;
  • trade-offs


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Forage fish play a pivotal role in marine ecosystems and economies worldwide by sustaining many predators and fisheries directly and indirectly. We estimate global forage fish contributions to marine ecosystems through a synthesis of 72 published Ecopath models from around the world. Three distinct contributions of forage fish were examined: (i) the ecological support service of forage fish to predators in marine ecosystems, (ii) the total catch and value of forage fisheries and (iii) the support service of forage fish to the catch and value of other commercially targeted predators. Forage fish use and value varied and exhibited patterns across latitudes and ecosystem types. Forage fish supported many kinds of predators, including fish, seabirds, marine mammals and squid. Overall, forage fish contribute a total of about $16.9 billion USD to global fisheries values annually, i.e. 20% of the global ex-vessel catch values of all marine fisheries combined. While the global catch value of forage fisheries was $5.6 billion, fisheries supported by forage fish were more than twice as valuable ($11.3 billion). These estimates provide important information for evaluating the trade-offs of various uses of forage fish across ecosystem types, latitudes and globally. We did not estimate a monetary value for supportive contributions of forage fish to recreational fisheries or to uses unrelated to fisheries, and thus the estimates of economic value reported herein understate the global value of forage fishes.

Introduction 2
Methods 3
Compilation and synthesis of Ecopath models3
Data extraction8
Importance of forage fish to ecosystem predators9
Direct and support service contributions of forage fish to commercial fisheries9
Forage fish contribution to global fisheries value10
Results 11
Quality of Ecopath models11
Extent of predator dependence on forage fish11
Support service contribution to ecosystem predator production13
Importance of forage fish to commercial fisheries13
Comparisons across latitude groups and ecosystem types15
Global estimate of forage fish value to fisheries15
Discussion 15
Acknowledgements 19
References 19
Supporting Information 22


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

‘Forage fish’ species are small or intermediate-sized pelagic species (e.g. sardine, anchovy, sprat, herring, capelin, krill) that are the primary food source for many marine predators, including mammals (Thompson et al. 1996; Pauly et al. 1998; Weise and Harvey 2008), seabirds (Crawford and Dyer 1995; Jahncke et al. 2004; Furness 2007; Daunt et al. 2008) and larger fish (Walter and Austin 2003; Butler et al. 2010; Logan et al. 2011; Magnussen 2011). Forage fish feed on zooplankton and phytoplankton and are important conduits of energy transfer in food webs for many marine ecosystems, from the tropics to the Earth's poles (Cury et al. 2000, 2003; Fréon et al. 2005; Bakun et al. 2010).

Fisheries for forage fish occur across broad latitudinal ranges (FAO 2010) and constitute a large and growing fraction of the global wild marine fish catch (Alder et al. 2008). In addition, five of the top ten fish species caught (by weight) in 2008 were forage fish species. Notably, the Peruvian anchoveta (Engraulis ringens, Engraulidae) supports the largest fishery in the world (FAO 2010). Nearly 90% of the global forage fish catch is used by reduction industries, which produce fish meal and fish oil (Alder et al. 2008). While economic studies of forage fish have focused primarily on their role as a directly harvested commodity (Herrick et al. 2009; Mullon et al. 2009; Tacon and Metian 2009), few have attempted to quantify the indirect economic contributions that these species provide (Hannesson et al. 2009; Herrick et al. 2009; Hannesson and Herrick 2010). Accounting for the indirect or support service values that prey species provide to other fisheries is inherently more difficult (Hannesson et al. 2009; Hannesson and Herrick 2010; Hunsicker et al. 2010), but doing so can provide important information to assess the trade-offs between exploiting forage fish and other species in the same marine ecosystem.

There has been growing scientific consensus for the application of ecosystem-based management approaches (Pikitch et al. 2004; McLeod et al. 2005; McLeod and Leslie 2009) in contrast to traditionally applied single-species approaches (Beddington et al. 2007; FAO 2010). Single-species management generally seeks to maintain populations of a target species yet ignores most ecosystem factors. Even in cases where forage fish are well managed from a single-species perspective (i.e. overfishing is not occurring), a form of ‘ecosystem overfishing’ sensu Murawski (2000) can occur, whereby depleted abundance of forage fish may negatively affect the ecosystem (Gislason 2003; Coll et al. 2008). Implementing an ecosystem-based approach to the management of forage fisheries seems especially warranted (Pikitch et al. 2004; Richerson et al. 2010; Smith et al. 2011), as these species exhibit strong trophic linkages and fluctuate in abundance along with seasonal, annual and inter-decadal variations in oceanographic forces (Barber and Chavez 1983; Francis et al. 1998; Polovina et al. 2001; Chavez et al. 2003).

Human decision-making is often influenced by comparisons of monetary values or trade-offs between different products or services (Polasky and Segerson 2009). By quantifying the value of these ecosystem products and services, such trade-offs, and the impacts of degrading ecosystems, are made more explicit (Costanza et al. 1997; Balmford et al. 2002; Barbier et al. 2011). The majority of economic analyses conducted for forage fish fisheries have been one dimensional (Herrick et al. 2009), focusing on factors or management strategies affecting the direct value of these species as a landed commodity. Only a handful of studies have enumerated the indirect values that species targeted by fisheries provide (Hannesson et al. 2009; Hannesson and Herrick 2010; Hunsicker et al. 2010; Kamimura et al. 2011). Because of their key position in marine food webs, the overall global importance of forage fish to fisheries and ecosystems has likely been significantly understated.

This study provides the first global estimate of forage fish value to commercially important marine fisheries and enumerates the contributions of forage fish to ecosystem predator production. We synthesized data obtained from Ecopath models representing marine ecosystems around the world. This approach allowed for broad relationships to be detected and described by summarizing data from multiple independent studies (Gurevitch and Hedges 1999), including information on feeding habits, production and catch rates. We estimated the contribution that forage fish species make to: (i) the diets and production of all forage fish predators within each modelled ecosystem, (ii) forage fish fisheries, in terms of catch and catch value and (iii) the catch and value of other commercially targeted predator species (e.g. tunas, cod, striped bass), based on their diet dependence on forage fish. We compared and contrasted these contributions and values, and investigated the effects of model structure, ecosystem type and latitude (Table 1). Finally, we use the relationships and properties revealed by these models, together with estimates of catch values at the scale of economic exclusive zones (EEZ) and high seas areas (HSA), to estimate the total value that forage fish contribute to global marine fisheries.

Table 1. List of the 72 Ecopath models used in this synthesis. Full model references can be found in Appendix S1 (available in the online version of this article)
Model No.Model nameModel year(s)Latitude groupEcosystem typeModel area (km2)Pedigree indexNo. of MGNo. of PMGNo. of FFMGValue dataKrill MGCitation
  1. 148 group model, 2pre-oil spill model, 3post-oil spill model, 4ETP7 model, 5La Niña model, 6El Niño model. MG = Model groups, PMG = Predator model groups, FFMG = Forage fish model groups.

1Western Bering Sea11980s–1990sHigh latitudeArctic high latitude254 00048223YesYesAydin et al. (2002)
2Eastern Bering Sea (1)1980sHigh latitudeArctic high latitude484 50825141YesNoTrites et al. (1999)
3Eastern Bering Sea (2)1980s–1990sHigh latitudeArctic high latitude485 00038192YesNoAydin et al. (2002)
4Prince William Sound, Alaska (1)21980–89High latitudeArctic high latitude88000.3511962YesNoDalsgaard and Pauly (1997)
5Prince William Sound, Alaska (2)31994–96High latitudeArctic high latitude90000.67548205YesNoOkey and Pauly (1999)
6Hecate Strait, Northern British Columbia2000High latitudeNon-upwelling coastal70 00050345YesYesAinsworth et al. (2002)
7Northern California Current1990UpwellingUpwelling69 17663383YesYesField et al. (2006)
8Gulf of California1978–79Tropical-SubtropicalSemi-enclosed27 9002581YesNoArreguín-Sánchez et al. (2002)
9Huizachi-Caimanero lagoon complex, Mexico1970–2000Tropical-SubtropicalTropical lagoon1750.7502661YesNoZetina-Rejón et al. (2003)
10Golfo de Nicoya, Costa Rica1980s–1990sTropical-SubtropicalTropical lagoon153020101YesNoWolff et al. (1998)
11Golfo Dulce, Costa Rica1960–90sTropical-SubtropicalTropical lagoon7502091NoNoWolff et al. (1996)
12Eastern Subtropical Pacific Ocean41993–97Tropical-SubtropicalOpen ocean32 800 00040312YesNoOlson and Watters (2003)
13Northern Humboldt Current51995–96UpwellingUpwelling165 0000.63832153YesNoTam et al. (2008)
14Northern Humboldt Current61997–98UpwellingUpwelling165 0000.63832163YesNoTam et al. (2008)
15Sechura Bay, Peru1996UpwellingUpwelling4000.4622251YesNoTaylor et al. (2008)
16Central Chile1998UpwellingUpwelling50 0422185YesYesNeira et al. (2004)
17Tongoy Bay, Chile1980s–1990sUpwellingUpwelling601751NoNoWolff (1994)
18Falkland Islands1990sTemperateNon-upwelling coastal527 00044322YesYesCheung and Pitcher (2005)
19South Brazil Bight1998–99Tropical-SubtropicalNon-upwelling coastal97 0002562YesNoGasalla and Rossi-Wongtschowski (2004)
20Caeté Estuary, Brazil1999Tropical-SubtropicalNon-upwelling coastal2201841YesNoWolff et al. (2000)
21Gulf of Paria1980s–1990sTropical-SubtropicalTropical lagoon760023111NoNoManickchand-Heileman et al. (2004)
22Northeastern Venezuela shelf1970s–1980sTropical-SubtropicalNon-upwelling coastal30 00016101YesNoMendoza (1993)
23Gulf of Salamanca1997Tropical-SubtropicalTropical lagoon9550.7431861YesNoDuarte and García (2004)
24Celestun lagoon, Mexico2001Tropical-SubtropicalTropical lagoon280.3621912YesNoVega-Cendejas and Arreguín-Sánchez (2001)
25Terminos lagoon, Mexico1980s–1990sTropical-SubtropicalTropical lagoon25002051NoNoManickchand-Heileman et al. (1998a)
26Southwestern Gulf of Mexico1980s–1990sTropical-SubtropicalTropical lagoon65 0001991NoNoManickchand-Heileman et al. (1998b)
27Laguna Alvarado, Mexico1991–94Tropical-SubtropicalTropical lagoon620.5003092YesNoCruz-Escalona et al. (2007)
28Tampamachoco lagoon, Mexico1980s–1990sTropical-SubtropicalTropical lagoon152361NoNoRosado-Solórzano and Guzmán del Próo (1998)
29Gulf of Mexico1950–2004Tropical-SubtropicalNon-upwelling coastal1 530 38761236NoNoWalters et al. (2008)
30West Florida shelf1980s–1990sTropical-SubtropicalNon-upwelling coastal170 00059182NoNoOkey et al. (2004)
31Chesapeake Bay2000TemperateNon-upwelling coastal10 0000.45045175YesNoChristensen et al. (2009)
32Gulf of Maine1977–86TemperateNon-upwelling coastal90 70030122YesNoHeymans (2001)
33Northern Gulf of St. Lawrence1985–87TemperateNon-upwelling coastal103 8120.65131193YesNoMorissette et al. (2003)
34Newfoundland1995TemperateNon-upwelling coastal495 0000.39650304YesNoHeymans and Pitcher (2002)
35Lancaster Sound region, Canada1980sHigh latitudeArctic high latitude97 6983221NoNoMohammed (2001)
36West Greenland1991–92High latitudeArctic high latitude63 5000.4391241YesNoPedersen (1994)
37Icelandic shelf1997High latitudeArctic high latitude115 0000.29521102NoNoMendy (1999)
38Barents Sea (1)1990High latitudeArctic high latitude1 400 00041185YesNoBlanchard et al. (2002)
39Barents Sea (2)1995High latitudeArctic high latitude1 400 00041185YesNoBlanchard et al. (2002)
40Baltic Sea1974–2000TemperateSemi-enclosed396 8381654YesNoHarvey et al. (2003)
41North Sea1981TemperateNon-upwelling coastal570 0002584YesYesChristensen (1995)
42English Channel1995TemperateNon-upwelling coastal89 60748154YesNoStanford and Pitcher (2004)
43Western English Channel1994TemperateNon-upwelling coastal56 45252204YesNoAraújo et al. (2005)
44Bay of Mont. St. Michel, France2003TemperateNon-upwelling coastal2501911YesNoArbach Leloup et al. (2008)
45Cantabrian Sea shelf1994TemperateNon-upwelling coastal16 0000.6692892YesNoSánchez and Olaso (2004)
46Azores Archipelago1997TemperateNon-upwelling coastal584 0000.40944151YesNoGuénette and Morato (2001)
47Northwestern Mediterranean Sea1994TemperateSemi-enclosed450023103YesNoColl et al. (2006)
48Orbetello lagoon, Italy1996TemperateNon-upwelling coastal27941YesNoBrando et al. (2004)
49Northern & Central Adriatic Sea1990sTemperateSemi-enclosed55 5000.65740163YesNoColl et al. (2007)
50Black Sea1989–91TemperateSemi-enclosed423 0001141YesNoÖrek (2000)
51Atlantic coast of Morroco1984UpwellingUpwelling586 9000.38238192YesNoStanford et al. (2004)
52Banc d'Arguin, Mauritanie1988–98Tropical-SubtropicalNon-upwelling coastal10 0000.5372271YesNoSidi and Diop (2004)
53Cape Verde Archipelago1981–85Tropical-SubtropicalNon-upwelling coastal53943191YesNoStobberup et al. (2004)
54Central Atlantic Ocean1997–98TemperateOpen ocean18 419 19139141YesNoVasconcellos and Watson (2004)
55Gambian continental shelf1995Tropical-SubtropicalNon-upwelling coastal40002372NoNoMendy (2004)
56Guinea-Bissau continental shelf1990–92Tropical-SubtropicalNon-upwelling coastal40 81632122YesNoAmorim et al. (2004)
57Senegambia1990Tropical-SubtropicalNon-upwelling coastal27 6001872YesNoSamb and Mendy (2004)
58Guinean continental shelf2005Tropical-SubtropicalNon-upwelling coastal42 96935212YesNoGascuel et al. (2009)
59Southern Benguela Current1990UpwellingUpwelling220 00032154YesNoShannon et al. (2003)
60South Orkneys/South Georgia1990sHigh latitudeAntarctic1 880 00030222YesYesBredesen (2004)
61Antarctic Peninsula1991–2001High latitudeAntarctic340039202YesYesErfan and Pitcher (2005)
62Kerguelen Archipelago EEZ1987–88TemperateNon-upwelling coastal575 10023152YesYesPruvost et al. (2005)
63Maputo Bay, Mozambique1980s–1990sTropical-SubtropicalNon-upwelling coastal11001041YesNoPaula e Silva et al. (1993)
64Great Barrier Reef, Australia2000Tropical-SubtropicalTropical lagoon325 84830122NoNoGribble (2005)
65Darwin Harbour, Australia1990–2000Tropical-SubtropicalNon-upwelling coastal2500.3752151NoNoMartin (2005)
66Brunei Darussalam1989–90Tropical-SubtropicalTropical lagoon73961341YesNoSilvestre et al. (1993)
67Terengganu, Malaysia1980sTropical-SubtropicalTropical lagoon10501322YesNoLiew and Chan (1987)
68Hong Kong, China1990sTropical-SubtropicalNon-upwelling coastal168037121NoNoBuchary et al. (2002)
69Tapong Bay, Taiwan1999–2001Tropical-SubtropicalTropical lagoon40.8201812NoNoLin et al. (2006)
70East China Sea1997–2000Tropical-SubtropicalOpen ocean770 0000.63645196YesNoJiang et al. (2008)
71Bohai Sea1982–83TemperateNon-upwelling coastal77 0001351YesNoTong et al. (2000)
72Central North Pacific1990sTropical-SubtropicalOpen ocean9 888 35025202NoNoCox et al. (2002)


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Compilation and synthesis of Ecopath models

Of the more than 200 Ecopath models that have been published (Fulton 2010), 72 were obtained and selected for this synthesis. The requirements for inclusion in our analysis were that the Ecopath models had to represent a marine or estuarine ecosystem in a relatively recent state (within the last 40 years), include at least one forage fish model group, and have all the necessary data and parameters openly available. The majority of Ecopath models used (90%, 65 out of 72) represented ecosystems within the past 30 years. We obtained Ecopath models from peer-reviewed publications (n = 33), technical reports (n = 36) and theses/dissertations (n = 3) (Table 1). Ecopath models that were not included failed to have at least one forage fish model group, did not have data openly available, represented older time periods (>40 years old), or a combination of all three. Collected models spanned a wide geographical range and provided relatively good global coverage of most coastal ocean areas and marine ecosystem types, with the exception of the Indian Ocean, which is poorly studied compared with other ocean areas (De Young 2006) (Fig. 1). When available, we also obtained Ecopath pedigree index information (Christensen and Walters 2004; Christensen et al. 2005) to assess data quality of the models.


Figure 1. Approximate locations of the 72 Ecopath models used in this synthesis. Ecopath models where monetary value could (white circles) and could not (black circles) be calculated. Model numbers correspond to Table 1.

Download figure to PowerPoint

To examine the patterns in forage fish contributions and values, we grouped Ecopath models by latitude and by ecosystem type. Latitude groupings consisted of three categories: Tropical-Subtropical (less than 30° N – less than 30° S), Temperate (greater than or equal to 30° N – 58° N and greater than or equal to 30° S – 58° S) and High latitude (greater than 58° N and greater than 58° S). We separated upwelling ecosystem models from the latitude groupings due to the dominant roles forage fish catches play in these ecosystems. Ecosystem types included: upwelling ecosystems, semi-enclosed ecosystems, non-upwelling coastal ecosystems, tropical lagoon ecosystems, open ocean ecosystems, Arctic high latitude ecosystems, and Antarctic ecosystems. All models were categorized into only one ecosystem type and latitude group (Table 1).

In this analysis we define ‘forage fish’ as species that occupy an important intermediary trophic position and that retain that ecological role throughout their life. We thus excluded from our definition species that assume this role early in life but later move into higher trophic categories as they age (e.g. North Pacific hake, Blue whiting, Alaska pollock).

Data extraction

We extracted model groups, catch data, diet composition matrices, biomass data, production-to-biomass ratios and model area (km2) from tables in Ecopath model publications and transferred them into separate Microsoft©Excel spreadsheets. When necessary, we converted all Ecopath catch and biomass data not conforming to the standard Ecopath units for catch (tonne km−2 year−1) and biomass (tonne km−2).

The majority (83%) of Ecopath models in this analysis had data on total catch (landings plus discards). The remaining 17% (12 out of 72) of the models only published landings data with no estimates of discards. For these 12 models we assumed that discards were zero in our analysis. Discards represent approximately 8% of the marine fisheries catch by weight globally but vary greatly among species and ecosystems (Kelleher 2005).

Ecopath models contain interactive ‘groups’ which can be composed of either single or multiple species that share similar life histories or ecological functions (Polovina 1984). We used the Ecopath models assembled with the original model groups as specified by the model authors. The published models generally included a list of species or taxa constituting each model group. When such taxonomic information was provided, we used this information to create an inventory of all species. In this study, we classified a model group as a forage fish group whenever at least one forage fish species was included. For instance, if an anchovy species was a component of a larger model group called ‘Small Pelagics’, along with gobies and juvenile mackerels, then we considered this group as a forage fish group, even though other species in that group may not necessarily meet our definition of forage fish. The majority (65% or 105 out of 161) of forage fish model groups consisted entirely of forage fish species. Of the remaining 56 forage fish model groups, 30 were discerned to be dominated by forage fish species, while information on the preponderance of forage fish species was lacking for the other 26 model groups. The one exception to our classification of forage fish model groups applied to krill (Order: Euphausiaea), which were only represented as separate model groups in 9 of the 72 Ecopath models in this analysis (Table 1). In the few remaining Ecopath models where krill were present in the ecosystem but not as a separate model group, they were grouped into various ‘Zooplankton’ groups. We chose to exclude these ‘Zooplankton’ model groups as forage fish groups in this analysis and only included contributions of krill from models with defined krill model groups. We acknowledge that this modelling approach may cause differences between ecosystems in terms of forage fish contributions (i.e. those that have a separate krill group and those that do not) but assumed in this analysis that if model authors grouped krill separately it was due to their perceived importance in the ecosystem. We considered it was more appropriate to include krill groups as forage fish in this analysis when present than to completely exclude them.

Importance of forage fish to ecosystem predators

We identified forage fish predators in all models and their dependence on forage fish (percent of forage fish in diet) from the respective model diet matrix. We defined forage fish predators as model groups whose diets contained any fraction of one or more forage fish model groups (i.e. diet of >0% forage fish). This definition allowed for forage fish species to be included as forage fish predators, if their diet consisted of forage fish. This rarely occurred, with only 3.9% (35 out of 895) of forage fish predators also included as forage fish. Forage fish predators were then categorized into the following dependence groups: (i) low dependence on forage fish (>0 to <25%), (ii) moderate dependence (≥25 to <50%), (iii) high dependence (≥50 to <75%) and (iv) extreme dependence on forage fish (≥75%).

We estimated the portion of each forage fish predator's production supported by forage fish across all ecosystem models using equations modified from Hunsicker et al. (2010). First, we calculated the total annual production (Pj, units: tonne km−2 year−1) of each forage fish predator group j in each Ecopath model using Equation (1), in which predator group j's biomass (Bj, units: tonne km−2) was multiplied by that respective predator group's production-to-biomass ratio (P B−1, units: year−1).

  • display math(1)

Second, we found the portion of each predator group's total annual production (Pi,j) supported by forage fish prey groups (i), by multiplying predator group j's respective diet dependence on forage fish (Di,j) by Pj using Equation (2).

  • display math(2)

The total support service contribution of forage fish to ecosystem predator production (Sz) therefore can be found using Equation (3), as the product of (Di,j) and (Pj) summed over all forage fish groups (i) and predator groups (j) in an ecosystem.

  • display math(3)

Hunsicker et al. (2010) showed that Di,j is equivalent to the contribution of prey group i to predator group j's production (Pi,j) when assimilation and energy content of prey items are roughly equivalent. In the absence of detailed data on these variables, we assumed they were equal to one another but note that our analysis underestimates Pi,j because of the generally high energy content of forage fish species (Van Pelt et al. 1997; Anthony et al. 2000) compared to most predators. Thus, our estimates for the support service contribution of forage fish to ecosystem predator production can be considered conservative in this regard.

Direct and support service contributions of forage fish to commercial fisheries

We calculated the contributions of forage fish to fisheries in terms of catch (tonne km−2 year−1) for all 72 Ecopath models and catch value (2006 USD km−2 year−1) for a subset of models that had adequate taxonomic information (n = 56). Ecopath models were grouped into categories based on ecosystem type and latitude of the model (Table 1). We used a global ex-vessel price database, developed by Sumaila et al. (2007) to obtain ex- vessel ‘real’ price data for all fished species in our Ecopath models. Ex-vessel ‘real’ price is defined as the actual prices that fishermen receive for their products before processing and is hereafter simply referred to as price. In this analysis, we use ‘value’ to refer to ex-vessel fish price times quantity (gross returns) and not economic profit (net returns).

We obtained total catch data for every country participating in fisheries in a respective Large Marine Ecosystem (LME) in year 2006 from the Sea Around Us project LME database (Watson et al. 2004;, and used the ex-vessel price database to compile country specific ex-vessel price data for every species in the 56 models. Information on every fishing country in each LME and their respective total catch can be accessed on the Sea Around Us project LME database website ( To account for differences in prices between countries operating in a given LME, we calculated a weighted average based on the total catch in 2006 of all participating countries within that LME. When model groups consisted of two or more species, the ex-vessel price for the model group was found by averaging the ex-vessel prices for all respective species within, which were each weighted by the catches of participating countries. We used these averaged ex-vessel model group prices to calculate fisheries value (2006 USD km−2 year−1) for each respective model group in all 56 Ecopath models.

For small geographic areas (e.g. estuaries, lagoons, and small coastal areas), we assumed that only the country surrounding these waters fished them. We made this assumption because detailed information about which specific countries fish within an Ecopath model area is not usually published. For the few Ecopath models that were located outside a defined LME area (e.g. Central North Pacific Ocean, Central Atlantic Ocean and Eastern Subtropical Pacific Ocean), we assumed participating fishing countries to be those nearest to, and surrounding, the model locations. Ecopath models of island countries and territories that fell outside of LME boundaries (e.g. the Azores Archipelago) were assumed to be fished only by that country, or the country of which it is a territory.

We estimated forage fish catch by summing the catch of all forage fish model groups in each respective ecosystem model. Catch value (2006 USD km−2 year−1) was estimated for each respective forage fish model group by multiplying the catch (tonne km−2 year−1) by the respective ex-vessel price (2006 USD tonne −1) (Sumaila et al. 2007). Similarly, we summed catch values for all forage fish model groups to find the total forage fish catch value (2006 USD km−2 year−1) for each Ecopath model. We estimated the support service contributions of forage fish to the catch (SC) and catch value (SV) of other commercially targeted model groups by using Equation (3), except that the predator group's total annual production (Pj) was replaced by the catch (Cj, Equation (4)) and catch value (Vj, Equation (5)) of each predator group j.

  • display math(4)
  • display math(5)

Forage fish contribution to global fisheries value

Forage fish species contribute to the value of global fisheries in two important ways: (i) by their direct catch value and (ii) by their support service as prey to the value of other commercially targeted species. Using forage fish value estimates for these contributions from each Ecopath model, we extrapolated to Exclusive Economic Zone (EEZ) or High Seas Area (HSA) regions to derive global estimates. We worked at the scale of EEZs and HSAs because independent estimates of forage fish catch values were available at this scale (Sumaila et al. 2007) to complement the values we estimated in Ecopath models. We assumed that a single Ecopath model representing an area within an EEZ or HSA region provided a reasonable depiction of the relationship between the support service value of forage fish and the total fisheries value for the entire region. A breakdown of the actual area covered by our Ecopath models as a percentage of the total EEZ/HSA area or the total Inshore fishing area (IFA) can be found in Table S1 (see Appendix S2). The IFA is defined by the Sea Around Us Project database ( as the area between the shoreline and whichever comes first, either the 200 m bathycline or a distance of 50 km from the shoreline. The majority of the global marine fisheries catch value (78%) and forage fish catch value (97%) is derived from IFAs (Sumaila et al. 2007)( A summary of Ecopath model coverage in terms of EEZ/HSA or IFA area and fisheries value is provided in Table S2 (see Appendix S2). When multiple Ecopath models were available for a given EEZ or HSA region, we used average values weighted by the geographic area covered by each ecosystem model. We quantified global forage fisheries value by summing the value of forage fish across all EEZs and HSAs in the Sea Around Us project database. The majority of forage fish species in these databases were separated into two commercial groups, ‘Herring-likes’ and ‘Anchovies’. We assumed that the total direct forage fish catch value for each respective EEZ and HSA was the sum of these two commercial groups. When data on ‘Herring-likes’ and ‘Anchovies’ were missing from this database, we used data available for forage fish categorized by species group. This method may slightly underestimate forage fisheries value, as it did not include some forage fish species that were grouped into other non-forage fish commercial groups.

To estimate the global support service value of forage fish to other commercially targeted species, we extrapolated the values estimated for each Ecopath model to each corresponding EEZ and HSA region. To do this, we used Ecopath models with value data available and calculated an Ecopath value ratio (EVR) using Equation (6). In Equation (6), the catch value of forage fish predators supported by forage fish (Sv) was divided by the total fishery catch value (y) of the Ecopath model, excluding non-cephalopod, non-krill invertebrates (e.g. other decapods, bivalves). By assuming that EVRs found in our Ecopath models are representative of the larger EEZs or HSAs in which they are located, we calculated the total support service value ($Supportive) of forage fish in each EEZ and HSA. Using Equation (7) we multiplied the respective EVR for an EEZ or HSA by the total fishery catch value (excluding non-cephalopod, non-krill invertebrates) for that area calculated from the Sea Around Us database ($SAUP).

  • display math(6)
  • display math(7)

Ecopath models were available for 25% (64 out of 257) of the world's EEZs and HSAs, which represents 33% of the total EEZ/HSA area (Table S2, Appendix S2). In the majority (36 out of 64) of these EEZ/HSA areas, Ecopath model coverage was >50% of the respective EEZ/HSA area (see Appendix S2, Tables S1 and S2). These EEZ/HSAs constitute 39% of the global marine catch value (2006 $USD) excluding non-cephalopod and non-krill invertebrates (i.e. other decapods, bivalves) and 53% of the global forage fish catch value (2006 $USD) (Table S2, Appendix S2). Ecopath model coverage of IFAs was even greater, representing 47% of the total area (km2) (Table S2, Appendix S2). An additional 86 EEZs and HSAs (see Table S1, Appendix S2), which did not have Ecopath models, were included under the assumption that the Ecopath model in the EEZ or HSA immediately adjacent was representative of that neighbouring EEZ or HSA. These EEZs and HSAs represented an additional 28% of the global forage fish catch value to fisheries. The remaining 107 EEZs or HSAs did not have Ecopath models or an adjacent neighbour with an Ecopath model (e.g. isolated islands) and represented only 19% of the global forage fish value to fisheries. In these EEZ/HSA areas, we applied an EVR based on the average of EVRs from other Ecopath models in the same latitudinal group. We calculated all values as ex-vessel price values in 2006 $USD and summed all support service values and forage fisheries catch values across all EEZs and HSAs. This produced our estimate of forage fish contribution to global fisheries value.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Quality of Ecopath models

Ecopath pedigree indices (Christensen and Walters 2004) were available for 22 models (Table 1). The Ecopath pedigree index varies with the quality of data within Ecopath models, and values can range from 0 (not reliable) to 1 (highly reliable) (Christensen and Walters 2004; Christensen et al. 2005). Ecopath pedigree indices in this analysis ranged from 0.295 to 0.820 with the majority (55%, 12 out of 22) exceeding 0.5 (Table 1). Differences were observed in pedigree indices of models published in peer-reviewed journals (Ecopath pedigree mean = 0.625, median = 0.638, n = 11) and technical reports (Ecopath pedigree mean = 0.450, median = 0.408, n = 11). None of our indices were in the poorest quality level grouping, wherein data are considered to be no better than guesses (<0.2; Christensen and Walters 2004; Christensen et al. 2005). Moreover, the average and median pedigree indices observed in this study (0.518 and 0.537, respectively) were substantially higher than those for other studies (0.441 and 0.439, respectively) (Morissette et al. 2006; Morissette 2007).

Extent of predator dependence on forage fish

Seventy-five percent (54 out of 72) of the Ecopath models used in this analysis had at least one model group that was highly (≥50% but <75% of diet) or extremely dependent (≥75% of diet) on forage fish. Twenty-nine percent (21 out of 72) of the models included at least one extremely dependent predator group. We found extremely dependent predators present across all latitude groups and ecosystem types, with the exception of open ocean ecosystems. Extremely dependent predators accounted for only 5.8% (52 out of 895) of all forage fish predators and consisted of fishes (n = 30), seabirds (n = 12), marine mammals (n = 9) and one species of squid (Loligo gahi, Loliginidae). Amongst conspecific predator groups, however, seabirds had the highest percentage of extremely dependent predators, with 19% (12 out of 62) of all seabird predators having diets ≥75% forage fish. Extremely dependent predators groups were most commonly found in upwelling and Antarctic ecosystem types, with an average of two and five extremely dependent predators per model, respectively. Many of these extremely dependent predator species were also listed on the IUCN Red List (Table 2).

Table 2. Extremely dependent forage fish predators (≥75% forage fish in their diets) found in this synthesis that have taxonomic information and are evaluated by the International Union for Conservation of Nature (IUCN) Red List. Model numbers correspond to model names in Table 1
Common nameScientific nameFamilyIUCN StatusaPopulation trendModel No(s)
  1. a

    IUCN (2011) IUCN Red List of Threatened Species. Version 2011.2 Downloaded on 2 December 2011.

Marine Mammals
Sei Whale Balaenoptera borealis BALAENOPTERIDAEEndangeredUnknown(1, 60)
Blue Whale Balaenoptera musculus BALAENOPTERIDAEEndangeredIncreasing(1, 60)
Fin Whale Balaenoptera physalus BALAENOPTERIDAEEndangeredUnknown(1, 60)
Common Minke Whale Balaenoptera acutorostrata BALAENOPTERIDAELeast ConcernStable(1, 60)
Southern Right Whale Eubalaena australis BALAENIDAELeast ConcernIncreasing60
Grey Seal Halichoerus grypus PHOCIDAELeast ConcernIncreasing40
Crabeater Seal Lobodon carcinophagus PHOCIDAELeast ConcernUnknown60
Humpback Whale Megaptera novaeangliae BALAENOPTERIDAELeast ConcernIncreasing(1, 60)
Ringed Seal Phoca hispida PHOCIDAELeast ConcernUnknown40
Black-browed Albatross Thalassarche melanophrys DIOMEDEIDAEEndangeredDecreasing18
Macaroni Penguin Eudyptes chrysolophus SPHENISCIDAEVulnerableDecreasing(60, 62)
Humboldt Penguin Speriscus humboldtii SPHENISCIDAEVulnerableDecreasing17
Peruvian Pelican Pelecanus thagus PELECANIDAENear ThreatenedDecreasing(13–14, 15, 17)
Guanay Cormorant Phalacrocorax bougainvillii PHALACROCORACIDAENear ThreatenedDecreasing(13–14, 15)
Sooty Shearwater Puffinus griseus PROCELLARIIDAENear ThreatenedDecreasing1
Gentoo Penguin Pygoscelis papua SPHENISCIDAENear ThreatenedDecreasing(60, 62)
King Penguin Aptenodytes patagonicus SPHENISCIDAELeast Concern62
Rhinoceros Auklet Cerorhinca monocerata ALCIDAELeast Concern1
Southern Rockhopper Penguin Eudypte schrysocome SPHENISCIDAELeast ConcernDecreasing62
Tufted Puffin Fratercula cirrhata ALCIDAELeast Concern1
Southern Giant-petrel Macronectes giganteus PROCELLARIIDAELeast ConcernDecreasing18
Cassin's Auklet Ptychoramphus aleuticus ALCIDAELeast Concern1
Peruvian Booby Sula variegate SULIDAELeast Concern(13–14, 15, 17)
Common Guillemot Uria aalge ALCIDAELeast Concern1
Yellowfin Tuna Thunnus albacares SCOMBRIDAENear ThreatenedDecreasing(13–14, 56)
Common Dolphinfish Coryphaena hippurus CORYPHAENIDAELeast ConcernStable(13–14)
West African Ladyfish Elops lacerta ELOPIDAELeast ConcernUnknown56
Skipjack Tuna Katsuwonus pelamis SCOMBRIDAELeast ConcernStable56
North Pacific Hake Merluccius productus MERLUCCIIDAELeast ConcernUnknown7
Sockeye Salmon Oncorhynchus nerka SALMONIDAELeast ConcernStable(4–5)
Pacific Bonito Sarda chiliensis SCOMBRIDAELeast ConcernDecreasing(13–14)

We evaluated the relative frequency of various levels of forage fish dependencies and how they varied across ecosystem types by combining data from all models. Pooled data across all ecosystem models indicated that on average, 49% of all predator groups in our models relied on forage fish for at least 10% of their dietary requirements (Fig. 2). Forage fish predators that are highly or extremely dependent on forage fish account for 16% of all predator groups in marine ecosystem models on average. Predators with diets consisting of more than 90% forage fish were also found but represented fewer than 5% of all predator groups in this analysis.


Figure 2. Percentage of forage fish predators in analysed ecosystems (n = 72) and their dependence on forage fish (% forage fish in diet). Solid line represents the Mean ± SD for all predators in this analysis. Ecosystem types: AA, Antarctic; OO, open ocean; U, upwelling current; HL, Arctic high latitude; SE, semi-enclosed; NUC, non-upwelling coastal; TL, tropical lagoon.

Download figure to PowerPoint

When comparing across ecosystem types, Antarctic ecosystem models generally had the greatest proportion of forage fish predators in their models for any level of forage fish dependence compared to other ecosystem model types (Fig. 2). Upwelling ecosystems had the second highest percentage of predators with 90% forage fish dependence levels. Tropical lagoon ecosystem types had the lowest proportion of predators for a given forage fish dependence level (Fig. 2).

Support service contribution to ecosystem predator production

The total predator production (tonne km−2 year−1) supported by forage fish varied greatly among the 72 models in this analysis (Fig. 3). Supported predator production was the largest for two upwelling ecosystem models, the northern California Current model and central Chile model, and one non-upwelling coastal ecosystem (Falkland Islands model). Forage fish contributed 52 and 17 tonne km−2 year−1 to predator production in northern California Current and central Chile models respectively, and the contribution in the Falkland Islands model was 18.9 tonne km−2 year−1. When the contribution of krill to the production of other forage fish (e.g. krill, sardines, anchovies) was removed in the northern California Current and Falkland Islands models, the support service to predators dropped to 32 and 3.3 tonne km−2 year−1 respectively.


Figure 3. Support service of forage fish to ecosystem predator production across all Ecopath models in this analysis (n = 72).

Download figure to PowerPoint

Across ecosystem types, the greatest support service contribution of forage fish to predator production was seen in upwelling and Antarctic ecosystems (Fig. 4a). The support service contribution to predator production in both these ecosystem types exceeded 9 tonne km−2 year−1, and were more than three times greater than values seen for Arctic ecosystems and non-upwelling coastal ecosystems and more than an order of magnitude greater than open-ocean, tropical lagoon and semi-enclosed ecosystem types (Fig. 4a). In terms of latitude groupings (with upwelling ecosystems excluded), we found the greatest support service contributions to predator production in high latitude regions (3.79 tonne km−2 year−1 ± 1.23 SE), followed by temperate latitudes (1.81 tonne km−2 year−1 ± 0.59 SE) and finally tropical-subtropical latitudes (1.18 tonne km−2 year−1 ± 0.17 SE; Fig. 4b).


Figure 4. Mean forage fish contribution to (non-commercial) ecosystem predator production by ecosystem type (a) and latitude grouping (b) with standard error plotted. Ecosystem types: U, upwelling current; TL, tropical lagoon; SE, semi-enclosed; OO, open ocean; NUC, non-upwelling coastal; HL, Arctic high latitude; and AA, Antarctic.

Download figure to PowerPoint

Importance of forage fish to commercial fisheries

Forage fish catch varied greatly among models examined, both in tonnage and ex-vessel price value. In some models, we found no forage fish catch reported (e.g. Central Atlantic Ocean), while others had extremely large forage fish catches (e.g. Sechura Bay, Peru). The highest forage fish catches were found in the Humboldt Current models where the Peruvian anchoveta fishery operates. Of the three Humboldt Current models, the Sechura Bay (Peru) model had an extraordinarily high level of forage fish catch (81 tonne km−2 year−1) valued at $35 497 (USD km−2 year−1), whereas in the northern Humboldt Current models for El Niño and La Niña periods, forage fish catches were 20 tonne km−2 year−1 ($934 USD km−2 year−1) and 39 tonne km−2 year−1 ($2020 USD km−2 year−1), respectively.

Forage fish contributed important support to other commercial fisheries in all models that contained such fisheries. Of the ecosystems we examined, forage fish were most important as prey, in terms of tonnage, to commercial fisheries in central Chile (3.82 tonne km−2 year−1), Prince William Sound (pre-oil spill model; 3.58 tonne km−2 year−1) and the northern California Current (3.13 tonne km−2 year−1; Fig. 5). In terms of value, forage fish provided the greatest support service to fisheries in the Prince William Sound model (pre-oil spill model) at a value of $5942 USD km−2 year−1, followed by the Chesapeake Bay at a value of $3095 USD km−2 year−1. The high support service values in these ecosystems are due to the large contribution of forage fish to the diets of salmon (Oncorhynchus spp., Salmonidae) in Prince William Sound and striped bass (Morone saxatilis, Percichthyidae) in Chesapeake Bay, both of which have relatively high ex-vessel price values.


Figure 5. Support service contributions of forage fish to other fisheries catch across all Ecopath models (n = 72).

Download figure to PowerPoint

In 13 out of 56 models, 100% of the total forage fish value was derived from support to other fisheries (i.e. there were no forage fish fisheries reported in these 13 ecosystems). In more than half the models (30 out of 56), the value of the fisheries supported by forage fish was greater than the value of forage fish catch (Fig. 6).


Figure 6. Percentage of total forage fish values (forage fish fisheries value + support service value to other fisheries) across Ecopath models (n = 56) derived from forage fish support service to other commercial fisheries. Ecosystems with 100% support service to other commercial fisheries do not have active forage fish fisheries in their respective ecosystem model.

Download figure to PowerPoint

Comparisons across latitude groups and ecosystem types

The largest forage fish catches were found in the tropical-subtropical latitude group (4.95 tonne km−2 year−1 ± 2.5 SE) and decreased monotonically as polar regions were approached. In contrast, the level of other commercial catch supported by forage fish was the lowest in the tropical-subtropical latitude group (0.23 tonne km−2 year−1 ± 0.05 SE) but greater in temperate (0.63 tonne km−2 year−1 ± 0.2 SE) and high latitude ecosystems (0.35 tonne km−2 year−1 ± 0.29 SE). We separated upwelling ecosystem models from these latitude groupings, as forage fish catches play a dominant role in these ecosystems. We found that temperate models had the highest forage fish fisheries catch when compared with the remaining two latitude groups (Fig. 7a). Forage fish catch value (excluding upwelling ecosystems) was the greatest in the tropical-subtropical latitude group and diminished poleward (Fig. 7b). The support service provided by forage fish for other commercial fisheries, in both catch and catch value, increased poleward so that it was equivalent (in catch) or exceeded (in catch value) the forage fish catch or catch value in high latitudes (Fig. 7a,b).


Figure 7. Mean catch (a) and mean catch value in 2006 USD (b) of forage fish (white bars) and mean supportive contribution of forage fish to other species' catch and catch value (grey bars), by latitude group. Bars indicate standard error. Upwelling ecosystem models were separated out to more clearly demonstrate latitudinal patterns.

Download figure to PowerPoint

Forage fish catch (tonne km−2 year−1) was the highest in upwelling ecosystems (Fig. 8a), exceeding that of all other ecosystem types combined by a factor of four. Forage fish catch exceeded the catch of other model groups that preyed on forage fish for all ecosystem types (Fig. 8a). Similarly, forage fish had the highest catch value in upwelling ecosystems at $5660 USD km−2 year−1 ± $4980 SE (Fig. 8b). Other ecosystem types had substantially lower forage fish catch values, each contributing <$830 USD km−2 year−1. The value of forage fish catches was the smallest in high latitude Arctic and Antarctic ecosystems ($184 USD km−2 year−1 and $149 USD km−2 year−1, respectively). In contrast, the support service value of forage fish was the greatest in the Arctic ecosystems (HL, mean = $706 USD km−2 year−1) – over 3.5 times greater than the forage fish value for that ecosystem type (Fig. 8b).


Figure 8. Mean catch (a) and catch value in 2006 USD (b) of forage fish (white bars) and mean supportive contribution of forage fish to other species' catch and catch value (grey bars). Bars indicate standard error. Ecosystem types: U, upwelling current; TL, tropical lagoon; SE, semi-enclosed; OO, Open ocean; NUC, non-upwelling coastal; HL, Arctic high latitude; and AA, Antarctic.

Download figure to PowerPoint

Global estimate of forage fish value to fisheries

The estimated total ex-vessel price value of forage fish to global commercial fisheries was $16.9 billion ($USD). This estimate combines global forage fish fishery value of $5.6 billion (33%, USD) with a support service value to other fisheries of $11.3 billion (67%, USD). This value represents nearly 20% ($16.9b/$85b) of the ex-vessel catch values of all world fisheries, estimated at between $80 and 85 billion USD year−1 (Sumaila et al. 2007; FAO 2010). Importantly, we found that the value of commercial fisheries supported by forage fish (e.g. cod, striped bass, salmon) was twice the value of forage fish fisheries at a global scale.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

We recognize that using Ecopath models, like any mathematical representation of an ecosystem, has certain limitations. However, our approach was built around the idea that, within the constraints of the model assumptions, averaging across many models will at least reduce the effects of stochastic uncertainty. Ecopath models provide only a single spatial and temporal representation of an ecosystem and they contain numerous assumptions whose consequences are often impossible to assess and could be important. This means, at the very least, that they do not capture changes in ecosystem dynamics and fisheries effort over space and time. Models are constructed based on data availability and the author's understanding of the ecosystem and research objectives, allowing for a gradient in model complexity and quality. The models contain simplified diet information of predators included in the models, which needs to be considered when interpreting or using the results of this study. For example, some Ecopath models lacked predators that are known to prey on forage fish, and in other cases, investigators pooled individual predator species together into a single trophic group. Nearly 30% (21 out of 72) of the models in our study did not have any seabird model groups, while 33% (24 out of 72) did not have a marine mammal group. Our estimates for predator production therefore are likely conservative, as we were not able to capture the importance of forage fish to these predators not included in the models. Likewise, aggregating predator species into model groups results in an averaged diet dependence on forage fish for the model group, which may mask high diet dependence for one or more individual species in that group. Averaging diet dependence for a single species over a large geographic area may also mask high diet dependencies that occur on smaller spatial or temporal scales. Validating every model to determine how well it represents its respective ecosystem and biological components was beyond the scope of this analysis, but Ecopath pedigree index information for a subset of models shows that the majority used in this analysis are of acceptable quality (Table 1). Using published models provided us with a large number of models covering the widest range of ecosystems and latitudes possible.

Here we used information on catches, catch values and food web connections to estimate the global contribution of forage fish to fisheries and ecosystems. While we find that the importance of forage fish varies geographically, it is clear that these species are of critical importance to many predators, including humans. We consider our approach as a reliable and relatively quick way of assessing the importance of forage fish in marine ecosystems and fisheries around the world. Ecopath models in this analysis covered 33% of the total EEZs and HSAs and covered 47% of the IFA (Table S2, Appendix S2), which is where 97% of the global forage fisheries catch value is derived (Sumaila et al. 2007). We acknowledge that geographic coverage is limited in the Indian Ocean. Although EEZ and HSA areas in the Indian Ocean account for 20% of the total EEZ and HSA area, they represent <15% of the total fisheries catch value (excluding non-cephalopod or non-krill invertebrates) and <12% of the total forage fish catch value. Furthermore, Indian Ocean EEZ and HSA areas accounted for <10% of the total global supportive value of forage fish. More robust fisheries information from this data-poor region (De Young 2006) would benefit future analyses.

At the global scale the supportive value of forage fish to fisheries greatly exceeds their direct commodity value. We note that the estimated total ex-vessel value ($16.9 billion USD annually) is likely an underestimate, because it does not take into account the contribution of forage species to early life history stages of predators that are not yet of commercial catch size (e.g. juvenile cod, juvenile striped bass). We also have not included in our analysis the contributions of species that are considered forage fish only during juvenile life stages (e.g. Alaska pollock). Accounting for these types of forage species would increase our estimates of support to ecosystem predator production and marine fisheries in certain ecosystems. More importantly, the ex-vessel value of commercial fisheries is only one of many other indicators of the economic contributions of forage fish, and thus is clearly an underestimate of total economic worth. We have not accounted for the potential economic value of forage fish to recreational fisheries, to ecotourism [e.g. the whale watching industry is estimated at $2.5 billion 2009 USD annually (Cisneros-Montemayor et al. 2010)], as bait for fisheries, and to the provision of other ecosystem services such as water filtration.

Forage fish are integral to marine food webs as prey for a wide variety of higher trophic-level species. For many predators, forage fish constitute a substantial percentage of their diet, possibly making them vulnerable to reductions or fluctuations in forage fish biomass. We found that many extremely dependent predators were species listed on the IUCN Red List as ‘Near Threatened’, ‘Vulnerable’ or ‘Endangered’ (Table 2). These predators were commonly found in upwelling ecosystems, where empirical evidence shows that changes in forage fish abundance – caused by fishing, the environment, or a combination of both – negatively impact predator reproduction (Sunada et al. 1981; Becker and Beissinger 2006), breeding (Crawford and Dyer 1995; Cury et al. 2011), abundance (Crawford and Jahncke 1999; Jahncke et al. 2004), and carrying capacity (Crawford et al. 2007). This analysis has identified ecosystems that are likely to have highly to extremely dependent forage fish predators and may assist in ecosystem-based management efforts that consider both commercial fisheries and effects on threatened or endangered species.

We provide the first global estimates of the importance of forage fish as support for predators in marine ecosystems. Quantifying forage fish catch, support service to other commercially targeted predators, and support to all other ecosystem predators allows for identification of potential trade-offs that may occur among uses (Fig. 9). Competition for the use of forage fish biomass among ecological and fisheries interests can result in trade-offs, which can lead to conflicts in the management of forage fish. This is especially important, as forage fish are an increasingly valued commodity (Naylor et al. 2009; Tacon and Metian 2009) and provide fundamental ecological support to many other species. Taking a holistic viewpoint of their value is a step towards quantification of the overall contributions forage fish make to marine ecosystems and to the global economy. A challenge that remains for fisheries managers and policy makers is determining acceptable levels of catch that account for the roles forage fish play in the larger marine environment.


Figure 9. Forage fish allocation across latitude groups in terms of support service to fisheries (grey bars), forage fish catch (white bars) and support service to ecosystem predator production (dotted grey bars).

Download figure to PowerPoint

The management of trade-offs in marine ecosystems can often be challenging (Okey and Wright 2004; Cheung and Sumaila 2008; Salomon et al. 2011), but accounting for trade-offs is important and can lead to more sustainable levels of exploitation without compromising ecosystem integrity (Okey and Wright 2004). Ultimately, accounting for trade-offs between forage fish fisheries and conservation goals will require knowledge and understanding of the sensitivity to which commercially targeted and non-commercial predator species respond to fisheries induced changes in forage fish abundance. A combination of modelling (Okey and Wright 2004; Cheung and Sumaila 2008; Smith et al. 2011) and empirical (Read and Brownstein 2003; Brodziak et al. 2004) methods will likely be required to fully understand trade-offs in forage fishery management.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

This work was supported by a grant from the Lenfest Ocean Program to E.K.P. and the research was conducted under the auspices of the Lenfest Forage Fish Task Force. This manuscript represents a portion of the Ph.D. dissertation research of K.J.R. at Stony Brook University. The authors would like to specifically thank the Sea Around Us project, a joint activity of the University of British Columbia and the Pew Environment Group, and L. Morissette for providing access to data essential to this work. Thanks also go to C. Perretti, S. Rahman and J. Rice for help in this analysis and fruitful discussions. Two anonymous reviewers are acknowledged and thanked for their comments which improved this manuscript.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Alder, J., Campbell, B., Karpouzi, V., Kaschner, K. and Pauly, D. (2008) Forage fish: from ecosystems to markets. Annual Review of Environment and Resources 33, 153166.
  • Anthony, J.A., Roby, D.D. and Turco, K.R. (2000) Lipid content and energy density of forage fishes from the northern Gulf of Alaska. Journal of Experimental Marine Biology and Ecology 248, 5378.
  • Bakun, A., Babcock, E.A., Lluch-Cota, S.E., Santora, C. and Salvadeo, C.J. (2010) Issues of ecosystem-based management of forage fisheries in “open” non-stationary ecosystems: the example of the sardine fishery in the Gulf of California. Reviews in Fish Biology and Fisheries 20, 929.
  • Balmford, A., Bruner, A., Cooper, P. et al. (2002) Economic reasons for conserving wild nature. Science 297, 950953.
  • Barber, R.T. and Chavez, F.P. (1983) Biological consequences of El Niño. Science 222, 12031210.
  • Barbier, E.B., Hacker, S.D., Kennedy, C., Koch, E.W., Stier, A.C. and Silliman, B.R. (2011) The value of estuarine and coastal ecosystem services. Ecological Monographs 81, 169193.
  • Becker, B.H. and Beissinger, S.R. (2006) Centennial decline in the trophic level of an endangered seabird after fisheries decline. Conservation Biology 20, 470479.
  • Beddington, J.R., Agnew, D.J. and Clark, C.W. (2007) Current problems in the management of marine fisheries. Science 316, 17131716.
  • Brodziak, J.K.T., Mace, P.M., Overholtz, W.J. and Rago, P.J. (2004) Ecosystem trade-offs in managing New England fisheries. Bulletin of Marine Science 74, 529548.
  • Butler, C.M., Rudershausen, P.J. and Buckel, J.A. (2010) Feeding ecology of Atlantic bluefin tuna (Thunnus thynnus) in North Carolina: diet, daily ration, and consumption of Atlantic menhaden (Brevoortia tyrannus). Fishery Bulletin 108, 5669.
  • Chavez, F.P., Ryan, J., Lluch-Cota, S.E. and Ñiquen, C.M. (2003) From anchovies to sardines and back: multidecadal change in the Pacific Ocean. Science 299, 217221.
  • Cheung, W.W.L. and Sumaila, U.R. (2008) Trade-offs between conservation and socio-economic objectives in managing a tropical marine ecosystem. Ecological Economics 66, 193210.
  • Christensen, V. and Walters, C.J. (2004) Ecopath with Ecosim: methods, capabilities and limitations. Ecological Modelling 172, 109139.
  • 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. (available online at
  • Cisneros-Montemayor, A.M., Sumaila, U.R., Kaschner, K. and Pauly, D. (2010) The global potential for whale watching. Marine Policy 34, 12731278.
  • Coll, M., Libralato, S., Tudela, S., Palomera, I. and Pranovi, F. (2008) Ecosystem overfishing in the Ocean. PLoS One 3, 110.
  • Costanza, R., d'Arge, R., de Groot, R. et al. (1997) The value of the world's ecosystem services and natural capital. Nature 387, 253260.
  • Crawford, R.J.M. and Dyer, B.M. (1995) Responses by 4 seabird species to a fluctuating availability of cape anchovy Engraulis capensis off South Africa. Ibis 137, 329339.
  • Crawford, R.J.M. and Jahncke, J. (1999) Comparison of trends in abundance of guano-producing seabirds in Peru and southern Africa. South African Journal of Marine Science 21, 145156.
  • Crawford, R.J.M., Underhill, L.G., Upfold, L. and Dyer, B.M. (2007) An altered carrying capacity of the Benguela upwelling ecosystem for African penguins (Spheniscus demersus). ICES Journal of Marine Science 64, 570576.
  • Cury, P., Bakun, A., Crawford, R.J.M. et al. (2000) Small pelagics in upwelling systems: patterns of interaction and structural changes in “wasp-waist” ecosystems. ICES Journal of Marine Science 57, 603618.
  • Cury, P., Shannon, L.J. and Shin, Y.-J. (2003) The functioning of marine ecosystems: a fisheries perspective. In: Responsible Fisheries in the Marine Ecosystem (eds M. Sinclair and G. Valdimarsson). CAB International, Wallingford, pp. 103123.
  • Cury, P.M., Boyd, I.L., Bonhommeau, S. et al. (2011) Global seabird response to forage fish depletion-One-third for the birds. Science 334, 17031706.
  • Daunt, F., Wanless, S., Greenstreet, S.P.R., Jensen, H., Hamer, K.C. and Harris, M.P. (2008) The impact of the sandeel fishery closure on seabird food consumption, distribution, and productivity in the northwestern North Sea. Canadian Journal of Fisheries and Aquatic Sciences 65, 362381.
  • De Young, C. (2006) Review of the State of World Marine Capture Fisheries Management: Indian Ocean. FAO Fisheries Technical Paper, Food and Agriculture Organization, Rome, p. 458.
  • FAO (2010) The State of World Fisheries and Aquaculture 2010. Food and Agriculture Organization, Rome, p. 197.
  • Francis, R.C., Hare, S.R., Hollowed, A.B. and Wooster, W.S. (1998) Effects of interdecadal climate variability on the oceanic ecosystems of the NE Pacific. Fisheries Oceanography 7, 121.
  • Fréon, P., Cury, P., Shannon, L. and Roy, C. (2005) Sustainable exploitation of small pelagic fish stocks challenged by environmental and ecosystem changes: a review. Bulletin of Marine Science 76, 385462.
  • Fulton, E.A. (2010) Approaches to end-to-end ecosystem models. Journal of Marine Systems 81, 171183.
  • Furness, R. (2007) Responses of seabirds to depletion of food fish stocks. Journal of Ornithology 148, 247252.
  • Gislason, H. (2003) The effects of fishing on non-target species and ecosystem structure and function. In: Responsible Fisheries in the Marine Ecosystem (eds M. Sinclair and G. Valdimarsson). CAB International, Wallingford, pp. 255275.
  • Gurevitch, J. and Hedges, L.V. (1999) Statistical issues in ecological meta-analyses. Ecology 80, 11421149.
  • Hannesson, R. and Herrick Jr, S.F. (2010) The value of Pacific sardine as forage fish. Marine Policy 34, 935942.
  • Hannesson, R., Herrick, S. and Field, J. (2009) Ecological and economic considerations in the conservation and management of the Pacific sardine (Sardinops sagax). Canadian Journal of Fisheries and Aquatic Sciences 66, 859868.
  • Herrick, S.F., Norton, J.G., Hannesson, R., Sumaila, U.R., Ahmed, M. and Pena-Torres, J. (2009) Global production and economics. In: Climate Change and Small Pelagic Fish (eds D.M. Checkley, J. Alheit, Y. Oozeki and C. Roy). Cambridge University Press, Cambridge, pp. 256274.
  • Hunsicker, M.E., Essington, T.E., Watson, R. and Sumaila, U.R. (2010) The contribution of cephalopods to global marine fisheries: can we have our squid and eat them too? Fish and Fisheries 11, 421438.
  • Jahncke, J., Checkley, D.M. and Hunt, G.L. (2004) Trends in carbon flux to seabirds in the Peruvian upwelling system: effects of wind and fisheries on population regulation. Fisheries Oceanography 13, 208223.
  • Kamimura, Y., Kasai, A. and Shoji, J. (2011) Production and prey source of juvenile black rockfish Sebastes cheni in a seagrass and macroalgal bed in the Seto Inland Sea, Japan: estimation of the economic value of a nursery. Aquatic Ecology 45, 367376.
  • Kelleher, K. (2005) Discards in the World's Marine Fisheries. An update. FAO Fisheries Technical Paper. No. 470. Food and Agriculture Organization of the United Nations (FAO), Rome, p. 131.
  • Logan, J.M., Rodriguez-Marin, E., Goni, N. et al. (2011) Diet of young Atlantic bluefin tuna (Thunnus thynnus) in eastern and western Atlantic foraging grounds. Marine Biology 158, 7385.
  • Magnussen, E. (2011) Food and feeding habits of cod (Gadus morhua) on the Faroe Bank. ICES Journal of Marine Science 68, 19091917.
  • McLeod, K.L. and Leslie, H. (2009) Ecosystem-based Management for the Oceans. Island Press, Washington, DC, USA, p. 368.
  • McLeod, K.L., Lubchenco, J., Palumbi, S.R. and Rosenberg, A.A. (2005) Scientific Consensus Statement on Marine Ecosystem-Based Managment. Signed by 217 academic scientists and policy experts with relevant expertise and published by the Communication Partnership for Science and the Sea at (last accessed on 13 November, 2011).
  • Morissette, L. (2007) Complexity, cost and quality of ecosystem models and their impact on resilience: a comparitive analysis, with emphasis on marine mammals and the Gulf of St. Lawrence. PhD thesis, University of British Columbia, Vancouver, BC, 260 pp.
  • Morissette, L., Hammill, M.O. and Savenkoff, C. (2006) The trophic role of marine mammals in the northern gulf of St. Lawrence. Marine Mammal Science 22, 74103.
  • Mullon, C., Mittaine, J.F., Thebaud, O., Peron, G., Merino, G. and Barange, M. (2009) Modeling the global fishmeal and fish oil markets. Natural Resource Modeling 22, 564609.
  • Murawski, S.A. (2000) Definitions of overfishing from an ecosystem perspective. ICES Journal of Marine Science 57, 649658.
  • Naylor, R.L., Hardy, R.W., Bureau, D.P. et al. (2009) Feeding aquaculture in an era of finite resources. Proceedings of the National Academy of Sciences of the United States of America 106, 1510315110.
  • Okey, T.A. and Wright, B.A. (2004) Toward ecosystem-based extraction policies for Prince William Sound, Alaska: integrating conflicting objectives and rebuilding pinnipeds. Bulletin of Marine Science 74, 727747.
  • Pauly, D., Trites, A.W., Capuli, E. and Christensen, V. (1998) Diet composition and trophic levels of marine mammals. ICES Journal of Marine Science 55, 467481.
  • Pikitch, E.K., Santora, C., Babcock, E.A. et al. (2004) Ecosystem-based fishery management. Science 305, 346347.
  • Polasky, S. and Segerson, K. (2009) Integrating ecology and economics in the study of ecosystem services: some lessons learned. Annual Review of Resource Economics 1, 409434.
  • Polovina, J.J. (1984) Model of a coral reef ecosystem: 1. The Ecopath model and its application to French Frigate Shoals. Coral Reefs 3, 111.
  • Polovina, J.J., Howell, E., Kobayashi, D.R. and Seki, M.P. (2001) The transition zone chlorophyll front, a dynamic global feature defining migration and forage habitat for marine resources. Progress in Oceanography 49, 469483.
  • Read, A.J. and Brownstein, C.R. (2003) Considering other consumers: fisheries, predators, and Atlantic herring in the Gulf of Maine. Conservation Ecology 7, 112.
  • Richerson, K., Levin, P.S. and Mangel, M. (2010) Accounting for indirect effects and non-commensurate values in ecosystem based fishery management (EBFM). Marine Policy 34, 114119.
  • Salomon, A.K., Gaichas, S.K., Jensen, O.P. et al. (2011) Bridging the divide between fisheries and marine conservation science. Bulletin of Marine Science 87, 251274.
  • Smith, A.D.M., Brown, C.J., Bulman, C.M. et al. (2011) Impacts of fishing low-trophic level species on marine ecosystems. Science 333, 11471150.
  • Sumaila, U., Marsden, A., Watson, R. and Pauly, D. (2007) A global ex-vessel fish price database: construction and applications. Journal of Bioeconomics 9, 3951.
  • Sunada, J.S., Kelly, P.R., Yamashita, I.S. and Gress, F. (1981) The brown pelican as a sampling instrument of age group structure in the northern anchovy population. California Cooperative Oceanic Fisheries Investigations Reports 22, 6568.
  • Tacon, A.G.J. and Metian, M. (2009) Fishing for feed or fishing for food: increasing global competition for small pelagic forage fish. Ambio 38, 294302.
  • Thompson, P.M., McConnell, B.J., Tollit, D.J., Mackay, A., Hunter, C. and Racey, P.A. (1996) Comparative distribution, movements and diet of harbour and grey seals from the Moray Firth, NE Scotland. Journal of Applied Ecology 33, 15721584.
  • Van Pelt, T.I., Piatt, J.F., Lance, B.K. and Roby, D.D. (1997) Proximate composition and energy density of some north pacific forage fishes. Comparative Biochemistry and Physiology Part A: Physiology 118, 13931398.
  • Walter, J.F. and Austin, H.M. (2003) Diet composition of large striped bass (Morone saxatilis) in Chesapeake Bay. Fishery Bulletin 101, 414423.
  • Watson, R., Kitchingman, A., Gelchu, A. and Pauly, D. (2004) Mapping global fisheries: sharpening our focus. Fish and Fisheries 5, 168177.
  • Weise, M.J. and Harvey, J.T. (2008) Temporal variability in ocean climate and California sea lion diet and biomass consumption: implications for fisheries management. Marine Ecology-Progress Series 373, 157172.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
faf12004-sup-0001-AppendixS1.docxWord document145KAppendix S1. Full references for Ecopath models used in this analysis in alphabetical order.
faf12004-sup-0002-AppendixS2.docxWord document178KAppendix S2. Supplementary Tables (Tables S1 and S2).

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.