seventh framework programme theme 7 environment ...eurobasin.dtuaqua.dk/eurobasin/documents... ·...

SEVENTH FRAMEWORK PROGRAMME THEME 7 Environment

Collaborative project (Large-scale Integrating Project) Project no: 246 933

Project Acronym: EURO-BASIN Project title: European Basin-scale Analysis, Synthesis and Integration

D5.4) Final report on top down trophic control of key pelagic species: on lower trophic level based in existing and knowledge assembled in EURO-BASIN [Month 46 -

Oct 2014; Resp: Geir Huse] Deliverable 5.4

Contributors: IMR, DTU Aqua, CLS, IFREMER, USTRATH, CEFAS, MRI-HAFRO, UHAM

Due date of deliverable: October 31, 2014 (month 46) Actual submission date: October 31, 2014.

Organisation name of the lead contractor of this deliverable: IMR Start date of project: 31.12.2010 Duration: 48 months

Project Coordinator: Michael St John, DTU Aqua

Project co-funded by the European Commission within the Seventh Framework Programme, Theme 6 Environment

Dissemination Level PU Public X PP Restricted to other programme participants (including the Commission) RE Restricted to a group specified by the consortium (including the Commission) CO Confidential, only for members of the consortium (including the Commission)

Deliverable 5.3 Preliminary progress report on top down trophic control of key pelagic species: on lower

trophic level based in existing and knowledge assembled in EURO-BASIN. Contribution to Task 5.2

Related Milestones: MS40, Joint workshop WP2-8.

Responsible: IMR Start month 1, end month 46

mailto:geir.huse(at)imr.no

EURO-BASIN | DX.X Deliverable NAME, AUTHOR, 2012

2

Contents Executive summary ........................................................................................................................... 3

Relevance to the project & potential policy impact ......................................................................... 3

Access to Data and/or model code ................................................................................................... 4

1 Diet and predators of small and medium sized pelagic fish species in the North Atlantic ........... 5

2 Stomach Contents of Herring, Blue Whiting, Mackerel, Albacore and Bluefin Tuna in the North Atlantic ..................................................................................................................................... 24

3 Analysis of stomach samples of small pelagic fish in the northern North Sea ............................ 35

4 Variability of albacore and bluefin tunas top-down trophic impacts in the Northeast Atlantic . 43

5 Spatially explicit estimates of prey consumption of the North Atlantic albacore Tuna (Thunnus alalunga) .................................................................................................................................. 46

6 Trophic impact of top predators migrations in exploited environments .................................... 49

7 Consumption of Calanus finmarchicus by planktivorous fish in the Norwegian Sea .................. 60

8 Effects of interactions between fish populations on ecosystem dynamics in the Norwegian Sea ................................................................................................................................................. 72

9 On growth variations in Northeast Atlantic blue whiting and resulting impacts on consumption ................................................................................................................................................. 82

10 Blue Whiting population modelling: progress, questions & next steps .................................... 97


3

Executive summary This report contains deliverable 5.4 of EUROBASIN - Final report on top down trophic control of key pelagic species: on lower trophic level based in existing and knowledge assembled in EURO-BASIN. It is a contribution to Task 5.2. The purpose of this task is to identify (size and composition) and quantify the trophic links between pelagic key species and their mid trophic level preys using existing information on prey abundance estimates, stomach content data analysis, and modelling. The aim is to calculate overall consumption by key species and determine the factors impacting this uptake (e.g. spatial distribution, physical conditions, competition) using data analysis and modelling. The work shows that there are regional differences in diet of the same species as well as differences between the species. Several models are presented below including a model based on the GADGET platform, the SEAPODYM, NORWECOM, as well as a size based model. The models can be used to generate estimates of consumption, distribution, and population dynamics and used to investigate effects of climate change and different harvest strategies on yield, stock stability and ecosystem state.

Top down effects by fish on zooplankton is for example addressed for the Norwegian Sea. In the Norwegian Sea there are rather strong overall spatial interactions within the pelagic stocks Norwegian spring spawing herring, blue whiting and mackerel during the feeding season, with considerable potential for exploitative competition for common zooplankton resources due to the large overlap in diet. Up until 2010 there was a strong build-up of biomass of planktivorous fish in the Norwegian Sea. The negative relationships between length at age and stock biomass, the pronounced reduction in zooplankton abundance witnessed in the Norwegian Sea during 2003-2009, the magnitude of consumption estimates and expansion in spatial distribution of fish indicate that the biomass of planktivorous fish in the area has been above the carrying capacity. All the stocks showed signs of density dependent length growth, whereas for herring and blue whiting there were also significant effects of interspecific competition. These results strongly indicates that top down regulation of zooplankton resources has been important in the Norwegian Sea in the last decade. In the most recent years there has been a reduction in the herring stock while the blue whiting and in particular the mackerel stock has increased. Concurrent with the decrease in the herring biomass, the zooplankton biomass has increased. Herring is generally the biggest zooplankton consumer in the Norwegian Sea. It can be argued that the herring has a greater predatory effect on the zooplankton since the herring enters the area in spring and feed on the the ascending generation of Calanus finmarchicus. before the main reproduction of Calanus has started. In this manner the Calanus is experiencing predation release with the reduction in the herring stock.

Regarding predation impacts by albacore and bluefin tunas, new dietary information has been collected for both species in the Bay of Biscay for several years during the 2000s showing regional and seasonal differences in diets. These data show that diets were dominated by a relatively small number of fish species, including Atlantic saury, blue whiting and anchovy. A new estimate of the predation impact by bluefin tuna during the period of historical fishery in the 1950s-early 1970s in the North Sea/Norwegian Sea has been obtained using newly-derived estimates of bluefin tuna biomass in these waters. These consumption estimates have been used together with a community size-spectrum analysis to derive a preliminary estimate of trophic cascade effects on lower trophic levels of the North Sea food web. These trophic modelling approaches will be continued in the next project period.

Relevance to the project & potential policy impact This report consists of 10 different chapters (see Contents) centered on top down trophic control of key pelagic fish species on lower trophic level. The species are mainly pelagic fish stocks including tuna, herring, blue whiting and mackerel and their consumption of zooplankton. The report includes both finished and ongoing work. The report focuses on commercially important stocks such as Norwegian spring spawning


4

herring, Northeast Atlantic mackerel, blue whiting and bluefin tuna. These findings therefore potentially have implications for management of these stocks.

Access to Data and/or model code A data publication in work package 1 (D1.10) are included in this deliverable entitled: " Stomach Contents of Herring, Blue Whiting, Mackerel, Albacore and Bluefin Tuna in the North Atlantic" by Pinnegar et al. (see below). The corresponding data has been delivered to Pangea for archiving (D1.8).


5

1 Diet and predators of small and medium sized pelagic fish species in the North Atlantic This section is part of a wider review on Comparative ecology of widely distributed pelagic fish species in the North Atlantic: implications for modelling climate and fisheries impacts

V. M. Trenkela, G. Huseb, B. R. MacKenziecd, P. Alvareze, H. Arrizabalagae, M. Castonguayf, N.

Goñie, F. Grégoiref, H. Hátúng, T. Jansend, J. A. Jacobseng, P. Lehodeyh, M. Lutcavagei, P. Marianid,

G. D. Melvinj, J. D. Neilsonj, L. Nøttestadb, G. J. Óskarssonk, M. R. Payned, D. E. Richardsonl, I.

Seninah, D.C. Speirsm a Ifremer, rue de l'île d'Yeu, BP 21105, 44311 Nantes cedex 3, France. b Institute of Marine Research -(IMR), Nordnesgate 33, 5817 Bergen, Norway. c Center for Macroecology, Evolution and Climate and Centre for Ocean Life, National Institute of Aquatic Resources (DTU Aqua), Charlottenlund Castle, 2920 Charlottenlund, Denmark. d Centre for Ocean Life, National Institute of Aquatic Resources (DTU Aqua), Charlottenlund

Castle, 2920 Charlottenlund, Denmark. e AZTI-Tecnalia, Herrera kaia portualdea z/g, 20110 Pasaia (Gipuzkoa), Spain. f DFO, Institut Maurice-Lamontagne, 850 route de la mer, C.P. 1000, Mont-Joli G5H 3Z4, Canada. g Faroe Marine Research Institute (FAMRI), FO-110 Tórshavn, Faroe Islands.

h CLS Satellite Oceanography Division, Ramonville St Agne, France. i LPRC, Umass Amherst, Marine Station, Box 3188, Gloucester, MA 01931, USA j DFO, Biological Station, 531 Brandy Cove Road, St. Andrews, E5B 2L9 Canada. k Marine Research Institute (MRI), Skulagata 4, 121, Reykjavik, Iceland. l NEFSC/NMFS/NOAA, 28 Tarzwell Drive, Narragansett, RI 02882, USA. m Department of Mathematics & Statistics, University of Strathclyde, Glasgow G1 1XH, UK. Introduction Here we review the available knowledge of the diet (prey) of each species, as well as their predators. We strive as much as possible to elucidate regional differences. The question asked is What are the commonalities and differences in their trophic roles? 1.1 Herring 1.1.1 Prey


6

Adult herring are opportunistic and feed on a variety of zooplankton and the larval stages of fish and invertebrates depending upon the location. However, throughout their range their primary food are copepods, such as Calanus and Pseudocalanus spp., and other small crustaceans (Scott and Scott, 1988; Prokopchuk and Sentyabov, 2006). In the Norwegian Sea, Calanus finmarchicus is the dominant zooplankton species (Wiborg, 1955), making up 80 % of the annual zooplankton production and the primary adult herring prey (Gislason and Astthorsson, 2002; Dommasnes et al., 2004; Prokopchuk and Sentyabov, 2006). NSSH follow C. finmarchicus through the Norwegian Sea, resulting in a general clockwise migration pattern during the feeding period. The quantity of C. finmarchicus that contributes to the total herring diet varies from 0 – 99 % depending on the temporal and spatial distribution of herring. Food composition of herring in the NW Atlantic varies in a similar way to the NE Atlantic, being dominated by one or two primary species, but including other organisms of appropriate size. The main prey of herring in Gulf of St Lawrence are also Calanus copepods, followed by capelin and euphausiids (Darbyson et al., 2003). It is not uncommon to find herring eggs in the stomachs of pre-spawning herring collected on spawning grounds in coastal and Southwest Nova Scotia (Gary Melvin, pers com.). The most important prey for herring collected on Georges Bank were chaetognaths, euphausiids, pteropods and copepods and in the Gulf of Maine it was euphausiids and copepods. In some areas herring have been found to feed on 0-group fish, including capelin, Sebastes spp. and herring themselves (Holst et al., 1997). Fish prey can even dominate the diet in some areas. Predation by juvenile NSSH in the Barents Sea is considered to impact year class strength of the local capelin stock, in addition to predation by 0-group cod and adult cod (Hjermann et al., 2010; Frank et al., 2011). On Georges Bank in the NW Atlantic, predation including that by herring, is believed to have contributed to the lack of a recovery of cod (Quinlan et al., 2000; Tsou and Collie, 2001; Murawski, 2010). In the North Sea where a more diverse group of prey organisms occur the principal herring prey are copepods (Calanus finmarchicus and Temora longicaudata), however, euphausids and post-larval fishes (Ammodytes spp. and clupeoids) and fish eggs (Pleuronectes platessa, and pelagic fishes) contribute also to their diet (Last, 1989; Segers et al., 2007). The summer of 2010 was anomalous with respect to weight-at-length, condition factor and fat content for a number of fish stocks on both sides of the north Atlantic (ICES, 2010b). Simultaneously, results from an international survey in the Nordic Seas in May indicated that zooplankton abundance had been declining, and in 2009-2010 it was at its lowest level since sampling started in 1997 (ICES, 2011b). Similar observations have also been reported for herring in the Gulf of Maine, Southwest Nova Scotia and the Gulf of St Lawrence where there is some evidence that the mean weight at age has been declining for several decades. Melvin and Martin (2012) found a significant relationship between mean monthly sea surface temperature, chlorophyll and herring body condition for specific months. They also noted that the decrease was not the same throughout the stock complex and varied among regions in the same stock. These observations could indicate a resource control on herring which would work via chlorophyll (plankton production), zooplankton through to fish growth on the western Atlantic. These observations evoke not only questions about trophic control but also about carrying capacity of the regional seas and gulfs. 1.1.2 Predators Herring are eaten by many predators at every stage from eggs to adult, and they are a key link in the transfer of energy from one trophic level to another in many ecosystems of the North


7

Atlantic. Eggs are preyed upon by numerous species of ground fish, invertebrates, and pelagic species, including herring themselves. As larvae they are consumed by fish and planktonic invertebrates, and filter feeding mammals. Once herring metamorphose into juveniles they become important prey for a wide variety of marine and terrestrial organisms, including seabirds which depend upon these small fish to feed the recently hatched chicks. As they increase in size so does the size of the predators feeding on them, and even the largest herring have several species that depend upon them for a major portion of their food consumption. Large predators of herring include seals, toothed whales (e.g. killer whale Orcinus orca), minke whale (Balaenoptera acutorostrata), fin whale (Balaenoptera physalus), humpback whale (Megaptera novaeangliae) and dolphin species, tuna (see below) and tuna like species, seabirds and various demersal fish species (Sigurjónsson and Víkingsson, 1997; Overholtz and Link, 2007). The total consumption of herring by predators is generally unknown and even when estimated it is usually only available for individual predator species or groups. Nevertheless, an estimate of total consumption of the Gulf of Maine-George Bank herring complex has been performed, and has shown that demersal fish species were the most effective predators, followed by marine mammals and large pelagic fish (Overholtz and Link, 2007). Changes in abundance of both prey and predators can cause major fluctuations in the estimate of total consumption (Overholtz et al., 2000). In the Norwegian Sea the predation pressure on NSSH during the feeding season is considered low as the whales focus more on zooplankton or capelin. Saithe (Pollachius virens) is known to prey on herring during the spawning period and they are believed to follow herring into the Norwegian Sea, but the extent of this is difficult to evaluate since there are very few samples of saithe from this area. In other areas of the north Atlantic, where herring aggregate for feeding, spawning, or overwintering, they are also followed by their predators (Parrish, 1993; Pitcher et. al., 1996; Read and Brownstein, 2003). It is not uncommon to observe whales, seals, seabirds, tuna, and a multitude of groundfish species feeding on herring spawning aggregations (Christensen, 1988; Purcell, 1990; Lindstrøm et al., 2000; Nøttestad, 2002; Overholtz et al. 2008). In some areas groundfish fisheries concentrate on herring spawning grounds to take advantage of their increased density and their eggs laying on the seafloor which attract demersal fish that feeds on them (Toresen, 1991; Livingston, 1993). 1.2 Mackerel 1.2.1 Prey Early life stages of Atlantic mackerel are characterized by fast growth and early feeding on copepod nauplii followed by a switch to piscivorous feeding habits at about 7 mm (Mendiola, et al., 2007; Robert et al., 2008). Early stages of mackerel exhibit selective feeding with calanoid copepods being preferred over cyclopoid copepods (Ringuette et al., 2002; Robert et al., 2008). When the larvae are > 6 mm and the potential growth rates are still increasing (Bartsch, 2002), high energy rich fish larvae become a central prey item. Piscivorous and cannibalistic feeding has been noted in all studies analysing mackerel larval feeding habits (Lebour, 1920; Grave, 1981; Ware and Lambert, 1985; Hillgruber and Kloppmann, 2001; Robert et al., 2008) with the exception of Last (1980), but this might be due to misidentification (Hillgruber and Kloppmann, 2001). Cannibalism was observed to be more prevalent at higher temperatures and increased with age and size (Mendiola et al., 2007). Comprehensive lists of prey species found in mackerel larvae stomachs are provided by Hillgruber and Kloppmann (2001), Robert et al. (2008), Hillgruber et al. (1997) and in references therein. Juvenile and adult Atlantic mackerel are opportunistic feeders that can ingest prey either by


8

particulate or filter feeding. They feed on a wide variety of zooplankton and small fish. Feeding behaviour and diet vary seasonally, diurnally, spatially and with size. Prey preference is positively size selective. Larger fish larvae are preferred over smaller larvae (Pepin et al., 1987; Pepin et al., 1988; Langøy, et al 2006; 2012). In the laboratory, prey size preference has been shown to be independent of prey concentration (Pepin et al., 1987; 1988). In the North Sea, the main zooplankton prey are copepods (mainly C. finmarchicus), euphausiids (mainly Meganyctiphanes norvegica) and hyperiids, while primary fish prey are sandeel, herring, sprat, and Norway pout. The most intensive period for mackerel feeding is April to August. Euphausiids are the main prey in winter and up to the beginning of spawning. Through spawning, summer and autumn, copepods and fish are also important parts of the diet (Mehl and Westgård, 1983; ICES, 1997). Mackerel in addition to herring is one of the major consumers of zooplankton in the Norwegian Sea, in particular of the dominant C. finmarchicus (Prokopchuk and Sentyabov, 2006; Langøy et al., 2012). Euphausiids and Themisto spp. also make up a significant bulk of the total zooplankton biomass in the Norwegian Sea (Dalpadado, 2002; Melle et al., 2004) and are among the preferred prey of mackerel (Langøy et al., 2012). The sea snail Limacina retroversa may also c o n t r i b u t e significantly to the diet in coastal Atlantic and Arctic water masses, even though more by weight than by numbers (Langøy et al., 2012). Mackerel has also been found to feed on adult capelin in frontal regions, illustrating their opportunistic and adaptive feeding behaviour (Nøttestad and Jacobsen, 2009). NWAM mackerel diet is dominated by copepods, decapods and fish larvae (Grégoire and Castonguay, 1989). Mackerel and herring are potential competitors in the Norwegian Sea both being opportunistic feeders with overlapping spatial distributions (Prokopchuk and Sentyabov, 2006). However, in some years (2004, 2006 and 2010) the degree of overlap in selection prey a n d distribution o f t h e s e t w o s p e c i e s h a s a p p e a r e d t o v a r y (Nøttestad et al., 2010; Utne et al., 2012b; Langøy et al., 2012). This perceived change could be due to stronger competition during the feeding season forcing the herring to the cooler fringe areas with poorer feeding. Support for this hypothesis is that herring were observed to be in poorer condition in 2010 than in previous years. 1.2.2 Predators A range of fish, mammal and bird predators have been reported to prey on mackerel (du Buit, 1996; Hunt and Furness, 1996; Overholtz et al., 2000; Olsen and Holst, 2001; Henderson and Dunne, 2002; Lewis et al., 2003; Trenkel et al., 2005). Locally mackerel can be important for some predators, such as killer whales in the northeast Atlantic and Norwegian Sea during summer (Nøttestad et al., Submitted). 1.3 Capelin 1.3.1 Prey Capelin is a planktivore with the main diet items being copepods, euphausiids and amphipods (see overview in Vilhjálmsson, 1994; Gjøsæter, 1998; Carscadden et al., 2001). Generally, the importance of copepods decreases with capelin size and that of euphausiids and amphipods increases. On the feeding grounds north of Iceland, euphausiids were estimated to constitute between 74-90% of the capelin diet (in weight), with corresponding estimates being somewhat lower for the Barents Sea (Vilhjálmsson, 1994). The importance of amphipods in the capelin diet is highest in the arctic waters where they are most abundant, for example in the northern Barents Sea (Gjøsæter, 1998) and the Labrador Sea (Carscadden et al., 2001). Capelin can


9

impact euphausiid density, as shown by an inverse relationship between their respective abundance estimates (Gjøsæter et al., 2002). Hassel et al. (1991) observed that the biomass of euphausiids in the upper water column was much lower in areas where capelin was present compared to where capelin was absent. Competition for food with other pelagic species is probably low for Icelandic capelin as they dominate the feeding grounds. In contrast, in the Barents Sea capelin may compete with polar cod (Boreogadus saida) in the eastern and northeastern part of the Barents Sea (Ushakov and Prozorkevich, 2002), and with juvenile herring in the southern parts (Huse and Toresen, 1996). In the Gulf of St. Lawrence in the NW Atlantic, interspecific feeding, competion of larvae of capelin, smelt (Osmerus mordax) and herring was considered minimal as they occurred in distinct water masses and had mainly different sizes (Courtois and Dodson, 1986). No information was found concerning competition with the adult part for the capelin stocks in the NW Atlantic. However, it can be expected that the more southerly and easterly distribution of the stocks since the early 1990s, and thereby a less offshore distribution, might have lead to increased competition with species normally occupying the continental shelves, such as herring. 1.3.2 Predators The large capelin stocks in the North Atlantic are important prey for a number of finfish, bird (Barrett et al., 2002; Carscadden et al., 2002), and marine mammal species (Carscadden et al., 2001; Dolgov, 2002). Gjøsæter (1998) considers capelin to play a key ecological role as an intermediary between zooplankton and higher tropic levels. Both cod and Greenland halibut (Reinhardtius hippoglossoides) feed heavily on capelin. The growth rates, somatic weight, and/or liver conditions of cod have been found to be positively related to biomass of capelin in the Barents Sea (Yaragina and Marshall, 2000), around Iceland (Vilhjálmsson, 2002) and in the NW Atlantic (Sherwood et al., 2007). Considering that capelin are an important forage species for many stocks, changes in their spatial distribution are likely to have significant consequences for their predators. For example, observed changes in capelin distribution, most likely caused by environmental factors, lead to them being less accessible to Greenland halibut (Dwyer et al., 2010), cod in the NW Atlantic (Rose and O‘Driscoll, 2002), and mature cod in Icelandic waters in the 2000s (Marine Research Institute, 2010). Capelin larvae are also heavily predated on. As mentionned above, predation by juvenile herring in the Barents Sea is considered to affect the year class strength of capelin (Hamre, 1994; Gjøsæter and Bogstad, 1998; Huse and Toresen, 2000); no information is available for predation on capelin larvae in Icelandic waters. The overlap between predators and juvenile capelin is usually higher than that for pre-spawning mature individuals which have a more oceanic distribution (Vilhjálmsson, 1994; Gjøsæter, 1998; Carscadden et al., 2001). 1.4 Blue whiting 1.4.1 Prey Blue whiting is a planktivorous species, with its dominant prey changing throughout lifetime. The diet of larval blue whiting consists predominately of Calanus spp, Pseudocalanus spp., Arcatia spp. and Oithona spp., with little or no phytoplankton or ichthyplankton (Conway, 1980). The diet of the juveniles and adults appears to be dominated by euphausiids together with Calanus spp.; small fish (Norway pout, pearlsides) also appear to be a part of the diet of the largest adults (Bailey, 1982; Bergstad, 1991; Dolgov et al., 2009). The abundance of all of these prey groups in the North Atlantic has been shown to have links to the sub-polar gyre (Hátún et al., 2009a) and


10

therefore changes in the available prey for blue whiting can be expected as a consequence of environmental change. As one of the major (by biomass) pelagic planktivorous species in the North Atlantic, it is almost inevitable that blue whiting competes with other species for resources. A recent study in the Barents Sea showed a high degree of dietary overlap between blue whiting and capelin, but also with herring and polar cod (Dolgov et al., 2009). It has recently been conjectured that the high abundance of pelagic fish in the Nordic Seas may be too large to be supported by the system‘s secondary (zooplankton) production (e.g. Payne et al 2012) although quantitative analyses capable of exploring this hypothesis in detail have yet to be carried out. 1.4.2 Predators Important predators of the southern component of blue whiting are hake in the Bay of Biscay (Guichet, 1995; Mahé et al., 2005) and hake, saithe, megrim, cod and whiting in the Celtic Sea (Pinnegar et al., 2003), in particularly during the summer months (Trenkel et al., 2005). Hake, saithe, and squid are potential predators in the northern regions (Bailey, 1982). Juvenile blue whiting have been identified as the main prey species of mackerel around the Iberian coast during autumn (Cabral and Murta, 2002; Olaso et al., 2005). Mackerel is hypothesised to be a major, and possibly controlling predator on juvenile blue whiting throughout its range (Payne et al., 2012). Several whale species also feed on blue whiting in the Bay of Biscay (Spitz et al., 2011), as does bluefin tuna. 1.5 Horse mackerel 1.5.1 Prey Horse mackerel is a planktivore, with the dominant prey being euphausiids and copepods, but also fish (Macer, 1977). Given its spatial overlap with other planktivores such as mackerel, blue whiting, and sardine, it is also likely to compete with these species for food, especially at an early age (Cabral and Murta, 2002). In the eastern part of the North Sea (off Jutland) horse mackerel were found to forage predominantly on fish (Dahl and Kirkegaard, 1987), with 0-group whiting being the most important prey, followed by other gadoids and herring. A shift in prey preference with age has been found: smaller individuals (< 20-24 cm) preyed mostly on crustaceans, gobies and haddock, while larger specimens shifted towards herring. For the Bay of Biscay, Letaconnoux (1951) and Olaso et al. (1999) provided a description of the horse mackerel diet. These observations indicated possible seasonal differences: during spring they preyed mainly on crustaceans, while in the autumn larger individuals (> 30 cm) began to prey on fish (blue whiting, gobiids, anchovy), which represented 45% of the food volume in this size-range. 1.5.2 Predators Horse mackerel is an important prey for cod, hake, megrim and whiting in the Celtic Sea, together with blue whiting (in summer) and mackerel (in winter) (Trenkel et al. 2005); it is also abundant in hake stomachs from the Bay of Biscay (Guichet, 1995; ) and those of a number of piscivores fish in the Cantabrian Sea (Preciado et al., 2008). In the Celtic Sea hake diet was found to reflect horse mackerel availability (Pinnegar et al., 2003), similarly in the Cantabrian Sea (Preciado et al., 2008). Horse mackerel are also consumed by bleufin tuna (see below). 2 Diet and predators of large pelagic fish species in the North Atlantic


11

2.1 Albacore 2.1.1 Prey Albacore is considered an opportunistic predator. In the Northeast Atlantic it has been reported to feed on fish, crustaceans and cephalopods with the most frequent and widespread prey being the euphausiid crustacean Meganyctiphanes norvegica (Aloncle and Delaporte, 1974; Ortiz de Zarate et al., 1987; Pusineri et al., 2005; Goñi et al. 2011). The most western investigations (up to 30ºW) were performed by Aloncle and Delaporte (1974) who found albacore fed mainly on amphipods (Themisto gaudichaudii), krill (Meganyciphanes norvegica) and the pelagic fish Cubiceps gracilis around the Azores islands. In the Bay of Biscay and surrounding areas, in addition to krill the sternoptychid fish Maurolicus muellerii as well as paralepidid fish represented an important part of albacore diet (Aloncle and Delaporte, 1974; Pusineri et al., 2005). However, as these species have become scarce in more recent years (Goñi et al., 2011), age-0 blue whiting has become a key prey, particularly along the shelf-break of the Bay of Biscay. Atlantic saury (Scomberesox saurus) have also been reported as an important prey for albacore in this zone in all studies to date. However, A t l a n t i c s a u r y is scarcer in the inner Bay of Biscay where sea surface temperature is higher (Aloncle and Delaporte, 1974). Anchovy is an important prey for albacore within the Bay of Biscay, mainly in the southern part (Ortiz de Zarate, 1987; Goñi et al., 2011), but is absent from the diet outside the bay. Average daily consumption of anchovy by albacore is around 10 individuals per day although a f t e r t h e r e c o v e r y o f t h e a n c h o v y s t o c k i n 2 0 1 0 a s m a n y a s 103 individuals h a v e b e e n f o u n d i n a n i n d i v i d u a l s t o m a c h (N Goñi, pers. comm.). The main spatial pattern in albacore diet is the difference between shelf-break areas and more oceanic areas with higher proportions of fish at the shelf break and more small crustaceans in oceanic waters (Goñi et al., 2011). In terms of feeding strategy, at the shelf- break albacore feed in the epipelagic layer during both daytime and night. In oceanic zones they feed in the epipelagic layer by night and dive into mesopelagic and/or bathypelagic layers to feed during the day (N Goñi pers comm.). These observations, together with the seasonal distribution of the fishing activity by surface gears (Sagarminaga and Arrizabalaga, 2010), suggest that the shelf-break areas are the main feeding areas for albacore in the NE Atlantic, whereas more oceanic areas would correspond to the last stages of the migration. Current albacore diet studies concern mainly juveniles, which compose the majority of albacore catches by surface fleets in the NE Atlantic. Their feeding ecology in the NW Atlantic has not been studied to date. 2.1.2 Predators Albacore is a top predator which probably has predators for juvenile stages. 2.2 Bluefin tuna 2.2.1 Prey Bluefin tuna in the north Atlantic consume a variety of fish species, as well as crustaceans and squid. Common fish prey species include herring, mackerel, anchovy, sardine, sprat, silver hake, squid, and demersal fish and invertebrate species, particularly in shallow continental regions (Chase, 2002; Rooker et al., 2007; Logan et al., 2011). Bluefin tuna in the North Sea and the Norwegian Sea consume herring, mackerel, sprat, garfish and gadoids (Tiews, 1978; Mather et al., 1995). Adult bluefin tuna in the Gulf of Maine primarily eat herring, sandlance and mackerel


12

(adults) (Crane, 1936; Chase, 2002; Estrada et al., 2005; Golet et al., 2007), while over a broad shelf area juvenile bluefin target sand lance (Chase, 2002; Golet et al., 2007; Logan et al., 2011). In the western Atlantic, stomach content analysis of long line caught bluefin and diving records from electronic tags show that in oceanic regions bluefin dive deeply and heavily target squids, with myctophids and other species identified to a lesser extent (Wilson et al, 1965; Logan et al., 2011). Stomach content studies of bluefin tuna captured south of Iceland in the late 1990s showed that diets in these waters were composed of European flying squid, Boreoatlantic gonate squid, barracudinas as well as a wide spectrum of other pelagic fish, squid and crustacean species (Olafsdottir et al. unpublished data, personal communication).The amount of prey consumed, and thus the predation impact that bluefin tuna formerly had in the North Sea has been estimated to be between 150-200,000 t of prey. Most (probably 75%) of this prey was herring (Tiews, 1978). This level of consumption occurred over a relatively short season because bluefin tuna were present in the North Sea for only 2-3 months per year (Tiews, 1978). The level of herring consumption by bluefin tuna in the 1950s was recently compared to consumption by all other predators (MacKenzie and Myers, 2007). The long-term mean consumption of herring by other predators was ca. 600,000 t during the mid-1960s-early 2000s (ICES, 2005). The bluefin consumption of herring could have been as high as 30% of that consumed by other predators, although in a much shorter period. This comparison suggests that predation by bluefin tuna on North Sea herring may have been quite substantial, and that bluefin tuna may therefore have been an important regulator of food web structure. The consumption of prey in the North Sea allowed bluefin tuna to increase their weights and condition factors before starting the return migration to southern waters in autumn (Tiews, 1978). Similar comparisons of predation impacts and condition have been conducted in the Gulf of Maine (Golet et al, 2007) and reveal r e l a t i o n s h i p s with prey availability, size, and energy status. In particular, significant associations between Atlantic bluefin tuna and Atlantic herring schools were identified (Golet et al., 2011), although long-term shifts in Atlantic herring distributions did not follow the same trend as for Atlantic bluefin tuna. The published dietary studies are mainly based on adult bluefins. Juvenile (ages 1-4) prey also on fish but also other lower trophic levels as revealed by both stomach content and isotopic analyses (Sara and Sara, 2007, Logan et al., 2011). Juvenile bluefin in the Bay of Biscay preyed primarily on 0-group anchovy, blue whiting, horse mackerel with myctophids, krill, swimming crabs and squid being consumed seasonally. Sandlance were the dominant prey species of juveniles in the Mid Atlantic Bight (Eggleston and Bochenek, 1990; Logan et al., 2011). This pattern is evident both in periods when sand lance was abundant and relatively rare. Other species of prey such as Atlantic mackerel, herring, butterfish and longfin squid were consumed in lesser amounts. In contrast, in the Bay of Biscay, consumption of anchovy seems to co-vary with local abundance, as consumption declined when the biomass of anchovy declined, and the consumption of some alternative prey species (e.g., krill) increased. Comparison of the estimated trophic levels of prey consumed based on prey remains in stomachs and isotopic measurements of bluefin tuna liver and muscle showed that trophic levels were lower based on isotopic evidence. Significant reduction in the condition of adult bluefin tuna in the Gulf of Maine has been linked to changes in the condition and availability of larger herring (Golet et al., 2007), possibly due to regional depletion, and bottom-up changes in trophic structure linked to oceanographic conditions (Golet et al., unpublished results). Prey switching is the norm for adult bluefin schools in the Gulf

of Maine, where schools travel up to 75 km d-1 and may switch feeding from sandlance to herring or other small pelagic species (Lutcavage et al., 2000; Gutenkunst et al., 2007).


13

2.2.2 Predators Bluefin tuna is a top predator but which has predators for juvenile stages. 2.3 Swordfish 2.3.1 Prey Swordfish as with tunas maintain their eyes and the central nervous system above ambient temperature, as a result having a vision system with high temporal resolution which is an advantage for hunting species (Fritches et al, 2005). The dominant prey swordfish consume are fish and cephalopods with crustaceans being a secondary prey type (Chancollon et al., 2006). Lanternfish, including Notoscopelus kroeyeri and Symbolophorus veranyi, are abundant, but paralepidids, Atlantic pomfret (Brama brama), and squid (Todarodes sagittatus, Ommastrephes bartramii, and Gonatus steenstrupi) dominate the diet by mass. Swordfish also appear to show feeding plasticity both among different areas and among animals in the same area. 2.3.2 Predators Larval swordfish are eaten by surface dwelling fishes, including larger swordfish (Scott and Scott, 1988). Yabe et al. (1959) described predation of young swordfish by blue sharks (Prionace glauca). As adults, swordfish have few natural enemies, but shortfin mako (Isurus oxyrinchus) sharks are frequently associated with attacks on hooked or harpooned swordfish (Scott and Scott, 1988). 2.4 Blue marlin 2.4.1 Prey Blue marlin are opportunistic feeders with substantial regional variation in their diets. For example the dominant prey items in blue marlin stomach contents were pomfret (Brama brama) and a squid (Ornithoteuthis antillarum) off Brazil (Junior et al., 2004), whereas skipjack (Katsuwonus pelamis) dominated in the western Pacific (Shimose et al., 2006), and frigate mackerel (Auxis thazard) in the Caribbean (Erdman, 2011). Analyses of food web structure consistently indicate that blue marlin is one of the top predators in pelagic ecosystems (Dambacher et al., 2010). In contrast to the adults, larval blue marlin are highly selective feeders. In the Straits of Florida, about 90% of stomach contents of small (<5mm) larval blue marlin were either a specific genus of copepod (Farranula) or a cladoceran (Evadne). The onset of piscivory occurred at 5 mm with exclusive piscivory occurring at 12 mm. Remarkably, despite the low productivity in the Straits of Florida relative to more temperate areas, blue marlin larvae had a high feeding incidence of 98% (Llopiz and Cowen, 2008). 2.4.2 Predators Blue marlin is a top predator which probably has predators for juvenile stages. 3 Discussion: Communalities and differences in the trophic roles of pelagic species across the North Altantic Based on the literature review, the trophic roles and controls of the studied pelagic species were determined (Table 1). In the trophic role classification, top-down effects of a species correspond to documented situations where abundance time-trends lead to detectable trends in the opposite directions in their preys, while for a bottom-up effect of a species similar time-trends in their predators have been found. If a pelagic species exerted both a top- down effect on its preys and a


14

bottom-up effect on its predators this suggests an overall middle-out food web control. Next, for all pelagic species we summarised the published evidence for their population abundance to be controlled by prey availability, called resource controlled, or predation pressure, called predator controlled. Due to its high abundance, broad distribution and position in the food web, herring is a key species in food webs throughout the North Atlantic. Herring are opportunistic zooplankton feeders focusing on the different dominant zooplankton species present. They also feed on ichthyoplankton and are cannibals (Holst, 1992). Mackerel larvae and juveniles are size- and species-selective feeders while adult mackerel are more opportunistic. Apart from disparate studies, the trophic role of blue whiting is poorly defined. However, mackerel, herring and blue whiting might be strong competitors in certain areas, such as the Norwegian Sea (Prokopchuk and Sentyabov, 2006; Langoy et al., 2012) where predation by these stocks appears to be responsible for the observed strong zooplankton reduction (Huse et al., 2012a). In this area there are indications that herring has been resource controlled in recent years via impacts on length growth (Huse et al., 2012a). In certain ecosystems and seasons, herring therefore exert a top-down effect on their prey, but in others they can exert a bottom-up effect on their predators (see tunas below). Mackerel could also exert a bottom-up effect on some of their predators (Nøttestad et al., Submitted). For blue whiting, extensive studies and stomach sampling programmes to examine both the predators and prey of this species would greatly improve our understanding of its dynamics, and its links to the environment and rest of the ecosystem. The currently available evidence points at a possible predation (on juveniles by mackerel, Payne et al., 2012) and competition control of the blue whiting population dynamics (e.g., by capelin, herring and polar cod, Dolgov et al., 2009). Capelin play an important role in local food webs, and both top-down and bottom-up effects have been observed (Skjoldal et al., 1992; Gjøsæter, 1998). They suffer predator control primarily via predation on their larvae. Finally, little is known about the food web role of horse mackerel in spite of it being a locally important prey. All investigated large pelagic species are more or less opportunistic feeders but due to their different spatial and vertical distributions their diet overlap is small. For example, while albacore is a nocturnal epipelagic feeder, swordfish feed during day and night in mesopelagic layers and hence the two species have different prey species and a very low trophic niche overlap (Pusineri et., 2008). It is currently unknown to what extent albacore might exert local top-down pressure. Given the observed flexibility in observed diet it seems unlikely that albacore are resource controlled. The extensive migrations of bluefin tuna for foraging imply that the predation impact by bluefin tuna on their prey populations is dispersed and seasonal; its magnitudes are not yet well documented but are probably modest to substantial. In contrast, there is evidence for local resource control of bluefin body condition. The diet of swordfish does not include any of the abundant small pelagic species considered here. There is no evidence for any type of food web effect by swordfish. Similarly for blue marlin, for which there is no published evidence for any type of food web effect in the North Atlantic. This may be due to a lack of studies in this area. In the eastern tropical Pacific Ocean Hunsicker et al. (2012) identified the potential for top-down control of sharks and billfishes on skipjack (Katsuwonus pelanis) and yellowfin tunas (Thunnus albacares).


15

Table 1. Summary of current knowledge of food web roles and interactions for selected small and large pelagic species in the North West and North East Atlantic (NEA). Major knowledge gaps are identified. Food web role/control of species: top-down effect of species on its prey; bottom-up effect of species on its predators; resource or predator impact on species population dynamics; competition; ? suspected. Species Stock structure Food web

role/control Differences between NE & NW Atlantic Important knowledge

gaps

herring several stocks in NE & NW Atlantic

top-down & bottom-up; resource controlled; competition with mackerel

oceanic feeding and overwintering only in NEA

Env. and stock size impact on migration, recruitment processes; top- down pressure.

mackerel

uncertain - probably weak structure in NE & NW Atlantic

bottom-up?; competition with herring

oceanic feeding only in NEA Env. and stock size impact on migration. Stock structure

capelin several populations in NE & NW Atlantic

top-down & bottom-up; predator controlled

higher fecundity at age/length in NW; NE deep-water and NW beach spawning

Recruitment processes, response to climate changes, food web role

blue whiting

uncertain in NE Atlantic

predator control; competition with capelin, herring?

mainly in NEA stock structure, food web role, dynamics of southern part/population

horse mackerel

several stocks in NE Atlantic

competition with mackerel, blue whiting, sardine?

only in NEA food web role

albacore

single population in N Atlantic

none single population food web impact

bluefin tuna

stocks in E & W Atlantic

top-down?, resource controlled?

maturation, abundance spawning areas, food web impact

sword-fish

possibly NE and NW population

none stronger effects of ocean currents on distribution in NWA

NE-NW Atlantic mixing uncertain

blue marlin

single population in Atlantic

none single population Migration patterns, spawning areas, juvenile distribution and ecology


16

4 Acknowledgements This review received funding from the European Union Seventh Framework Programme project EURO-BASIN (ENV.2010.2.2.1-1) under grant agreement n° 264933. 5 References Aloncle, H., Delaporte, F., 1974. Données nouvelles sur le germon Atlantique Thunnus alalunga

Bonnaterre 1788 dans le Nord-Est Atlantique. 1ère Partie – Rythmes alimentaires et circadiens. Revue des Travaux de l’Institut des Pêches Maritimes 37, 475-572.

Bailey, R., 1982. The population biology of blue whiting in the North Atlantic. Advances in Marine Biology 19, 257–355.

Barrett, R.T., Asker-Nilsen, T., Gabrielsen, G.W., and Chapdelaine, G., 2002. Food consumption by seabirds in Norwegian waters. ICES Journal of Marine Science 59, 43–57.

Bartsch, J., 2002. Modelling the temperature mediation of growth in larval fish. Fisheries Oceanography 11, 310-314.

Bergstad, O.A., 1991. Distribution and trophic ecology of some gadoid fish of the Norwegian Deep: 1 Accounts of individual species. Sarsia 75, 269-313.

Cabral, H.N., Murta, G., 2002. The diet of blue whiting, hake, horse mackerel and mackerel off Portugal. Journal of Applied Ichthyology 18, 14-23.

Carscadden, J.E., Frank, K.T., 2002. Temporal variability in the condition factors of Newfoundland capelin (Mallotus villosus) during the past two decades. ICES Journal of Marine Science 59, 950-958.

Carscadden, J.E., Frank, K.T., Leggett, W.C., 2001. Ecosystem changes and the effects on capelin (Mallotus villosus), a major forage species. Canadian Journal of Fisheries and Aquatic Sciences 58, 73-85.

Chancollon, O., Pusineri, C., Ridoux, V., 2006. Food and feeding ecology of Northeast Atlantic swordfish (Xiphias gladius) off the Bay of Biscay. ICES Journal of Marine Science 63, 1075-1085.

Chase, B.C., 2002. Differences in diet of Atlantic tuna (Thunnus thynnus) at five seasonal feeding grounds of the New England continental shelf. Fisheries Bulletin 100: 168-180.

Christensen, I., 1988. Distribution, movements and abundance of killer whales (Orcinus orca) in Norwegian coastal waters, 1982-1987, based on questionnaire surveys. In: Sigurjónsson and Leatherwood (eds) North Atlantic killer whales. Rit fiskideildar 11, 79-88.

Conway, D.V.P., 1980. The food of larval blue whiting, Micromesistius poutassou (Risso), in the Rockall area. Journal of Fish Biology 16, 709-723.

Courtois, R., Dodson, J.J., 1986. Diet and principal factors influencing the feeding of the larvae of capelin (Mallotus villosus), of smelt (Osmerus mordax) and of herring (Clupea harengus harengus) in a partially mixed estuary. Canadian Journal of Fisheries and Aquatic Science 43, 968-979.

Crane, J., 1936. Notes on the biology and ecology of giant tuna Thunnus thynnus, L., observed at Portland, Maine. Zoologica 212, 207-212.

Dahl, K., Kirkegaard, E., 1987. Stomach contents of mackerel, horse mackerel and whiting in the eastern part of the North Sea in July 1985. ICES C.M. 1986/H: 68 12 pp.

Dalpadado, P., 2002. Inter-specific variations in distribution, abundance and possible life-cycle patterns of Themisto spp. (amphipoda) in the Barents Sea. Polar Biology 25, 656-666.

Dambacher, J.M., Young, J.W., Olson, R.J., Allain, V., Galván-Magaña, F., Lansdell, M.J., Bocanegra-


17

Castillo, N., Alatorre-Ramírez, V., Cooper, S.P., Duffy, L.M., 2010. Analyzing pelagic food webs leading to top predators in the Pacific Ocean: A graph-theoretic approach. Progress in Oceanography 86, 152-165.

Darbyson, E., Swain, D.P., Chabot, D., Castonguay, M. 2003. Diel variation in feeding rate and prey composition of herring and mackerel in the southern Gulf of St Lawrence. Journal of Fish Biology 63, 1235-1257.

Dolgov, A.V. 2002. The role of capelin (Mallotus villosus) in the foodweb of the Barents Sea. ICES Journal of Marine Science, 59: 1034–1045.

Dolgov, A.V., Johannesen, E., Heino, M., Olsen, E., 2009. Trophic ecology of blue whiting in the Barents Sea. ICES Journal of Marine Science 67, 483-493.

Dommasnes, A., Melle, W., Dalpadado, P., Ellertsen, B. 2004. Herring as a major consumer in the Norwegian Sea. ICES Journal of Marine Science 61, 739-751.

du Buit, M.H., 1996. Diet of hake (Merluccius merluccius) in the Celtic Sea. Fisheries Research 28, 381-394.

Dwyer, K.S., Buren, A., Koen-Alonso, M., 2010. Greenland halibut diet in the Northwest Atlantic from 1978 to 2003 as an indicator of ecosystem change. Journal of Sea Research, 64, 436-445.

Eggleston, D.B., Bochenek, E.A., 1990. Stomach contents and parasite infestation of school bluefin tuna Thunnus thynnus collected from the Middle Atlantic Bight, Virginia. Fisheries Bulletin 88, 389-395.

Erdman, D.S., 2011. The sport fishery for blue marlin off Puerto Rico. Transactions of the American Fisheries Society, 91, 225-227.

Estrada, J.A., Lutcavage, M., Thorrold, S., 2005. Diet and trophic position inferred from stable carbon and nitrogen isotopes of Atlantic bluefin tuna (Thunnus thynnus). Marine Biology 147: 37-45.

Frank, K.T., Petrie, B., Fisher, J.A.D., Leggett, W.C., 2011. Transient dynamics of an altered large marine ecosystem. Nature. doi:10.1038/nature10285. 6p.

Fritches, K.A., Brill, R.W., Warrant, E.J., 2005. Warm eyes provide superior vision in swordfish. Current Biology 15: 55-58.

Fromentin, J.-M., 2009. Lessons from the past: investigating historical data from bluefin tuna fisheries. Fish and Fisheries 10, 197-216.

Gislason, A., Astthorsson, A.S. 2002. The food of Norwegian spring-spawning herring in the western Norwegian Sea in relation to the annual cycle of zooplankton. Sarsia 87, 236-247.

Gjøsæter, H., 1998. The population biology and exploitation of capelin (Mallotus villosus) in the Barents Sea. Sarsia, 83, 453-496.

Gjøsæter, H., Bogstad, B., 1998. Effects of the presence of herring (Clupea harengus) on the stock-recruitment relationship of Barents Sea capelin (Mallotus villosus). Fisheries Research 38, 57-71.

Gjøsæter, H., Dalpadado, P., Hassel, A., 2002. Growth of Barents Sea capelin (Mallotus villosus) in relation to zooplankton abundance. ICES Journal of Marine Science 59, 959-967.

Golet, W.J., Cooper, A.B., Campbell, R., Lutcavage, M., 2007. Decline in condition of Atlantic bluefin tuna in the Gulf of Maine. Fisheries Bulletin 104, 390-395.

Biscay. In: Nielsen, J., Arrizabalaga, H., Fragoso, N., Hobday, A., Lutcavage, M., Sibert, J. (eds) Tagging and Tracking of Marine Animals with Electronic Devices pp 51-73.

Goñi, N., Logan, J., Arrizabalaga, H., Jarry, M., Lutcavage, M., 2011. Variability of albacore (Thunnus alalunga) diet in the North-East Atlantic and Mediterranean Sea. Marine Biology 158, 1057-1072.


18

Grave, H., 1981. Food and feeding of mackerel larvae and early juveniles in the North Sea. Rapp. P. -v. Reun. Const. Int. Explor. Mer. 178, 454-459.

Grégoire, F., M. Castonguay. 1989. L'alimentation du maquereau bleu (Scomber scombrus) dans le golfe du Saint-Laurent et sur le plateau néo-écossais, avec une application du test de Mantel. Can. Tech. Rep. Fish. Aquat. Sci. 1673: vi + 23.

Guichet, R., 1995. The diet of European hake (Merluccius merluccius) in the northern part of the Bay of Biscay. ICES Journal of Marine Science 52, 21-31.

Gutenkunst, R., Newlands, N., Lutcavage, M., Keshett, L., 2007. Inferring resource distributions from Atlantic bluefin tuna movements: an analysis based on net displacement and length of track. Journal of Theoretical Ecology 245, 243-257.

Hamre, J., 1994. Biodiversity and the exploitation of the main fish stocks in the Norwegian-Barents Sea ecosystem. Biodiversity Conservation 3, 473-492.

Hassel, A., Skjoldal, H.R., Gjøsæter, H., Loeng, H., Omli, L., 1991. Impact of grazing from capelin (Mallotus villosus) on zooplankton: a case study in the northern Barents Sea in August 1985. In: Proceedings of the Pro Mare Symposium on Polar Marine Biology, Trondheim, Norway, 12–16 May 1990. Edited by Sakshaug, E., C.C.E. Hopkins, and N.A. Oritsland. Polar Res. 10: 371–388.

Hátún, H., Payne, M.R., Beaugrand, G., Reid, P.C., Sandø, A.B., Drange, H., Hansen, B., Jacobsen, J.A., Bloch, D., 2009a. Large bio-geographical shifts in the north-eastern Atlantic Ocean: From the subpolar gyre, via plankton, to blue whiting and pilot whales. Progress In Oceanography 80, 149-162.

Henderson, A.C., Dunne, J., 2002. Stomach contents of spiny dogfish Squalus acanthias L. off the west coast of Ireland. Ir. Nat. J. 3, 101-105.

Hillgruber, N., Kloppmann, M., 2001. Small-scale patterns in distribution and feeding of Atlantic mackerel (Scomber scombrus L.) larvae in the Celtic Sea with special regard to intra-cohort cannibalism. Helgoland Marine Research 55, 135-149.

Hillgruber, N., Kloppmann, M., Wahl, E., Von Westernhagen, H., 1997. Feeding of larval blue whiting and atlantic mackerel: A comparison of foraging strategies. Journal of Fish Biology 51, 230-249.

Hjermann, D.Ø., Bogstad, B., Dingsør, G.E., Gjøsæter, H., Ottersen, G., Eikeset, A.M., Stenseth, N.C., 2010. Trophic interactions affecting a key ecosystem component: a multistage analysis of the recruitment of the Barents Sea capelin (Mallotus villosus). Canadian Journal of Fisheries and Aquatic Science 67, 1363-1375.

Holst, J.C., Iversen, S.A., 1992. Distribution of Norwegian spring-spawning herring and mackerel in the Norwegian sea in late summer, 1991. ICES CM 1992/H:13.

Holst, J.C., Salvanes, A.G.V., Johansen, T., 1997. Feeding, Ichthyophonus sp. infection, distribution and growth history of Norwegian spring-spawning herring in summer. Journal of Fish Biology 50, 652-664.

Hunsicker, M.E., Olson, R.J., Essington, T.E., Maunder, M.N., Duffy, L.M., Kitchell, J.F., 2012. Potential top-down control on trophical tunas based on sized-based structure of predator-prey interactions. Marine Ecology Progress Series 445, 263-277.

Huse, G., Holst, J.C., Utne, K., Nottestad, L., Melle, W., Slotte, A., Ottersen, G., Fenchel, T., Uiblein, F., 2012. Effects of interactions between fish populations on ecosystem dynamics in the Norwegian Sea - results of the INFERNO project Preface. Marine Biology Research 8, 415-419.

Huse, G., Toresen, R., 2000. Juvenile herring prey on Barents Sea capelin larvae. Sarsia 85, 385-391. Huse, G., Toresen, R. 1996. A comparative study of the feeding habits of herring (Clupea harengus,


19

Clupeidae, L) and Capelin (Mallotus villosus, Osmeridae, Muller) in the Barents Sea. Sarsia 81, 143-153.

Hunt, G.J., Furness, R.W., 1996. Seabird/fish interactions, with particular reference to seabirds in the north sea. ICES cooperative research report. Nov 1996.

ICES, 2011b. Report of the Working Group on Widely Distributed Stocks (WGWIDE). ICES CM 2011/ACOM:15. 624 pp.

ICES, 2010b. Report of the ICES Advisory Committee, 2010 Book 9. 299pp. ICES, 1997. Database report of the stomach sampling project 1991. ICES cooperative research

report 219. Junior, T., Vooren, C.M., Lessa, R.P., 2004. Feeding Habits of Four Species of Istiophoridae (Pisces:

Perciformes) from Northeastern Brazil. Environmental Biology of Fishes 70, 293-304. Langøy, H., Nøttestad, L., Skaret, G., Broms, C., Fernö, A., 2012. Overlap in distribution and diets of

Atlantic mackerel (Scomber scombrus), Norwegian spring-spawning herring (Clupea harengus) and blue whiting (Micromesistius poutassou) in the Norwegian Sea during late summer. Marine Biology Research, 8, 442-460.

Langøy, H., Nøttestad, L., Skaret, G., Cecilie, T., Broms, A., Fernö, A., 2006. Feeding ecology of Atlantic mackerel (Scomber scombrus) in the Norwegian sea: Diet, prey selection and possible food competition with herring (Clupea harengus) in different water masses. ICES CM 2006/ F:12.

Last, J.M., 1989. The food of herring, Clupea harengus, in the North Sea, 1983–1986. Journal of Fish Biology 34, 489-501.

Lebour, M.V. ,1920. The food of young fish III. Journal of the Marine Biological Association UK 12, 261-324.

Letaconnoux, R., 1951. Contribution a l'étude des espéces du genre Trachurus et spécialment du Trachurus trachurus. Mém. Off. Scient. Peches Marit. 15, 1-67.

Lewis, S., Sherratt, T.N., Hamer, K.C., Harris, M.P., Wanless, S., 2003. Contrasting diet quality of northern gannets Morus bassanus at two colonies. Ardea 91, 167-176.

Lindstrøm, U., Harbitz, A., Haug, T., Pedersen, T., 2000. Foraging behaviour of minke whales (Balaenoptera acutorostrata) in the southern Barents Sea. ICES CM 2000/Q:07, 16p.

Livingston, P.A., 1993. Importance of predation by groundfish, marine mammals and birds on walleye pollock theragra chalcogramma and pacific herring clupea pallasi in the eastern bering sea. Marine Ecology Progress Series 102, 205-215.

Llopiz, J.K., Cowen, R.K., 2008. Precocious, selective and successful feeding of larval billfishes in the oceanic Straits of Florida. Marine Ecology Progress Series 358, 231-244.

Logan, J.M., Rodriguez-Marin, Goñi, N. Barreiro, S., Arrizabalaga, H., Golet, W., Lutcavage, M., 2011. Diet of young Atlantic bluefin tuna (Thunnus thynnus) in eastern and western Atlantic forage grounds. Marine Biology 158, 73-85.

Lutcavage, M.E, Brill, R.W., Goldstein, J.L., Skomal, G.B., Chase, B.C., Tutein, J., 2000. Movements and behavior of adult North Atlantic bluefin tuna (Thunnus thynnus) in the northwest Atlantic determined using ultrasonic telemetry. Marine Biology 137, 347-358.

Macer, C.T. 1977. Some aspects of the biology of the horse mackerel (Trachurus trachurus (L.)) in waters around Britain. Journal of Fish Biology 10, 51-62.

Mahé, K., Amara, R., Bryckaert, T., Kacher, M., Brylinski, J.M., 2007. Ontogenetic and spatial variation in the diet of hake (Merluccius merluccius) in the Bay of Biscay and the Celtic Sea. ICES Journal of Marine Science 64, 1210-1219.

Marine Research Institute, 2010. State of marine stock in Icelandic waters 2009/2010; prospects for


20

the quota year 2010/2011. Hafrannsóknir, No. 153. 178 pp. Mather, F.J. III, Mason Jr., J.M., Jones, A.C., 1995. Historical document: Life history and fisheries of

Atlantic bluefin tuna. NOAA Technical Memorandum NMFS-SEFSC-370. Mehl, S., Westgård, T., 1983. The diet and consumption of mackerel in the North Sea (a preliminary

report). ICES CM 1983/H:34. Melle, W., Ellertsen, B., Skjoldal, H.R., 2004. Zooplankton: The link to higher trophic levels. In: The

Norwegian Sea Ecosystem (Skjoldal, H.R., ed.) Tapir Academic Press, pp. 137-202. Mendiola, D., Alvarez, P., Cotano, U., De Murguia, A.M., 2007. Early development and growth of the

laboratory reared north-east Atlantic mackerel Scomber scombrus L. Journal of Fish Biology 70, 911-933.

Melvin, G.D., Martin, R. 2012. Review of environmental factors, condition factor, and fat content of the 4WX herring. Canadian Science Advisory Secretariat Research Document 2012/XXX. 26p.

Murawski, S.A., 2010. Rebuilding depleted fish stocks: the good, the bad, and mostly, the ugly. ICES Journal of Marine Science 67, 1830–1840.

Nøttestad, L. 2002. Killer whales (Orcinus orca L.) and saithe (Pollachius virens L.) trap herring (Clupea harengus L.) in shallow water by taking advantage of steep bottom topography. ICES CM 2002/N:20. 11p

Nøttestad, L., Jacobsen, J.A., 2009. Coordinated norwegian-faroese ecosystem survey with M/V ”Libas”, M/V ”Eros”, and M/V “Finnur fríði” in the Norwegian sea, 15 july- 6 august 2009. ICES CM 2009/ACOM:12:WD.

Nøttestad, L., Jacobsen, J.A., Sveinbjørnsson et al. 2010. Cruise report from the coordinated ecosystem survey with M/V "libas" and M/V "brennholm", M/V "finnur fridi" and R/V "arni fridriksson" in the Norwegian sea and surrounding waters, 9 july- 20 august 2010.

Olafsdottir, D., Ingimundardottir, T., 2000. Preliminary report on experimental fisheries and biological research on bluefin tuna (Thunnus thynnus) within the Iceland EEZ, 1996-1998. Collective volume of scientific papers ICCAT, Madrid, Vol. 91: 827-838.

Olaso, I., Cendrero, O., Abaunza, P., 1999. The diet of the horse mackerel, Trachurus trachurus (Linnaeus 1758), in the Cantabrian Sea (north of Spain). Journal of Applied Ichthyology 15, 193-198.

Olaso, I., Gutiérrez, J.L., Villamor, B., Carrera, P., Valdés, L., Abaunza, P., 2005. Seasonal changes in the north-eastern Atlantic mackerel diet (Scomber scombrus) in the north of Spain (ICES Division VIIIc). Journal of the Marine Biological Association of the UK 85, 415-418.

Olsen, E., Holst, J.C., 2001. A note on common minke whale (Balaenoptera acutorostrata). Journal of Cetacean Research and Management, 2, 179-183.

Orlova, E.L., Rudneva, G.B., Renaud, P.E., Eiane, K., Savinov, V., Yurko, A.S., 2010. Climate impacts on feeding and condition of capelin Mallotus villosus in the Barents Sea: evidence and mechanisms from a 30 year data set. Aquatic Biology 10, 105-118.

Ortiz de Zarate, V., Cort, J.L. 1998. Albacore (Thunnus alalunga, Bonnaterre) stock structure in the Atlantic Ocean, as inferred from distribution and migration patterns. Proceedings of the ICCAT Tuna Symposium (Beckett, J.E. ed.), Vol. 1, 251-260.

Overholtz, W.J., Jacobson, L.D., Link, J.S., 2008. An Ecosystem Approach for Assessment Advice and Biological Reference Points for the Gulf of Maine-Georges Bank Atlantic Herring Complex. North American Journal of Fisheries Management 28, 247-257.

Overholtz, W.J., Link, J.S., 2007. Consumption impacts by marine mammals, fish, and seabirds on the Gulf of Maine-Georges Bank Atlantic herring (Clupea harengus) complex during the years 1977-2002. ICES Journal of Marine Science 64, 83-96.


21

Overholtz, W.J., Link, J.S., Suslowicz, L.E., 2000. Consumption of important pelagic fish and squid by predatory fish in the northeastern USA shelf ecosystem with some fishery comparisons. ICES Journal of Marine Science 4, 1147-1159.

Parrish, J.K., 1993. Comparison of the hunting behavior of four piscine predators attacking schooling prey. Ethology 95, 233-246.

Payne, M.R., Egan, A., Fässler, S.M.M., Hátún, H., Holst, J.C., Jacobsen, J.A., Slotte, A., Loeng, H., 2012. The rise and fall of the NE Atlantic blue whiting (Micromesistus poutassou). Marine Biology Research, 8, 475-487.

Pepin, P., Koslow, J.A., Pearre, S., 1988. Laboratory study of foraging by atlantic mackerel, Scomber scombrus, on natural zooplankton assemblages. Canadian Journal of Fisheries and Aquatic Sciences 45, 879-887.

Pepin, P., Pearre, S., Koslow, J.A., 1987. Predation on larval fish by atlantic mackerel, Scomber scomburs, with a comparison of predation by zooplankton. Canadian Journal of Fisheries and Aquatic Sciences 44, 2012-2017.

Pinnegar, J.K., Trenkel, V.M., Tidd, A.N., Dawson, W.A., du Buit, M.-H., 2003. Does diet in Celtic Sea fishes reflet prey availability? Journal of Fish Biology 63 (Supplement A), 197-212.

Pitcher, T.J., Misund, O.A., Fernö, A., Totland, B., Melle, V., 1996. Adaptive behaviour of herring schools in the Norwegian Sea as revealed by high-resolution sonar. ICES Journal of Marine Science, 53: 449–452.

Preciado, I., Velasco, F., Olaso, I., 2008. The role of pelagic fish as forage for the demersal fish community in the southern Bay of Biscay. Journal of Marine Systems 72, 407–417.

Prokopchuk, I., Sentyabov, E., 2006. Diets of herring, mackerel, and blue whiting in the Norwegian Sea in relation to Calanus finmarchicus distribution and temperature conditions. ICES Journal of Marine Science 63, 117-127.

Purcell, J.E., 1990. Soft-bodied zooplankton predators and competitors of larval herring (Clupea harengus pallasi) at herring spawning grounds in british columbia. Canandian Journal of Fisheries and Aquatic Sciences 47, 505-515.

Pusineri, C., Chancollon, O., Ringelstein, J., Ridoux, V. 2008. Feeding niche segregation among the Northeast Atlantic community of oceanic top predators. Marine Ecology Progress Series 361, 21-34.

Pusineri, C., Vasseur, Y., Hassani, S., Meynier, L., Spitz, J., Ridoux, V., 2005. Food and feeding ecology of juvenile albacore, Thunnus alalunga, off the Bay of Biscay: a case study. ICES Journal of Marine Science 62,116–122.

Quinlan, J.A., Lough, R.G., Michaels, W., Fogarty, M., Buckley, L.J., Manning, J.P., Durbin, E., Runge, J.A., Werner, F.E., 2000. Examining the potential effects of vertebrate predation on Georges Bank larval cod: a modeling study for the 1995 field season. ICES CM 2000/N:25.

Read, A.J., Brownstein, C.R., 2003. Considering other consumers: fisheries, predators, and Atlantic herring in the Gulf of Maine. Conservation Ecology 7(1): 2. [online] URL: http://www.consecol.org/vol7/iss1/art2/

Reid, P.C., Borges, M.F., Svendsen, E., 2001. A regime shift in the North Sea circa 1988 linked Ringuette, M., Castonguay, M., Runge, J.A., Grégoire, F., 2002. Atlantic mackerel (Scomber

scombrus) recruitment fluctuations in relation to copepod production and juvenile growth. Canadian Journal of Fisheries and Aquatic Sciences 4, 646-656.

Robert, D., Castonguay, M., Fortier, L., 2008. Effects of intra- and inter-annual variability in prey field on the feeding selectivity of larval Atlantic mackerel (Scomber scombrus). Journal of Plankton Research 30, 673-688.

http://www.consecol.org/vol7/iss1/art2/















22

Rooker, J.R., Alvarado Bremer, J.R., Block, B.A., Dewar, H., De Metrio, G., Corriero, A., Kraus, R.T., Prince, E.D., Rodriguez-Marin, E., Secor, D.H., 2007. Life history and stock structure of Atlantic bluefin tuna (Thunnus thynnus). Reviews in Fisheries Science 15, 265–310.

Rose, G.A., O’Driscoll, R.L., 2002. Capelin are good for cod: can the northern stock rebuild without them? ICES Journal of Marine Science 59, 1018-1026.

Sagarminaga, Y, Arrizabalaga, H., 2010. Spatio-temporal distribution of albacore (Thunnus alalunga) catches in the northeastern Atlantic: relationship with the thermal environment. Fisheries Oceanography 19, 121–134.

Sara, G., Sara, R., 2007. Feeding habits and trophic levels of bluefin tuna Thunnus thynnus of different size classes in the Mediterranean Sea. Journal of Applied Ichthyology 23, 122–127.

Scott, W.B., Scott, M.G., 1988. Atlantic fishes of Canada. University of Toronto Press, Toronto. 608 pp.

Segers, F.H.I.D, Dickey-Collas, M., Rijnsdorp, A.D., 2007. Prey selection by North Sea herring (Clupea harengus), with special reference to fish eggs. ICES Journal of Marine Science 64, 60–68.

Sherwood, G.D., Rideout, R.M., Fudge, S.B., Rose, G.A., 2007. Influence of diet on growth, condition and reproductive capacity in Newfoundland and Labrador cod (Gadus morhua): Insights from stable carbon isotopes ( 13C). Deep Sea Research II 54: 2794-2809.

Shimose, T., Shono, H., Yokawa, K., Saito, H., Tachihara, K., 2006. Food and feeding habits of blue marlin, Makaira nigricans, around Yonaguni Island, southwestern Japan. Bulletin of Marine Science, 79, 761-775.

Sigurjónsson, J., 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.

Skjoldal, H.R., Gjøsæter, H., Loeng, H., 1992. The Barents Sea ecosystem in the 1980s: ocean climate, plankton and capelin growth. ICES Marine Science Symposium 195, 278-290.

Spitz, J., Cherel, Y. Bertin, S., Kiszka, J., Dewez, A., Ridoux, V., 2011. Prey preferences among the community of deep-diving odontocetes from the Bay of Biscay, Northeast Atlantic. Deep-Sea Research Part 1: Oceanographic Research Papers 58, 273-282.

Tiews, K., 1978. On the disappearance of bluefin tuna in the North Sea and its ecological implications for herring and mackerel. Rapp. P.-v. Reun. Cons. int. Explor. Mer. 172, 301-309.

Toresen, R., 1991. Predation on the eggs of norwegian spring-spawning herring (Clupea harengus L.) on a spawning ground on the west coast of Norway. ICES Journal of Marine Science 48, 15-21.

Trenkel, V.M., Pinnegar, J.K., Dawson, W.A., du Buit, M.H., Tidd, A.N., 2005. Spatial and temporal structure of predator–prey relationships in the Celtic Sea fish community. Marine Ecology Progress Series 299, 257–268.

Tsou, T.S., Collie, J.S., 2001. Predation-mediated recruitment in the Georges Bank fish community. ICES Journal of Marine Science 58, 994–1001.

Ushakov, N.G., Prozorkevich, D.V. 2002. The Barents Sea capelin – a review of trophic interrelations and fisheries. ICES Journal of Marine Science, 59: 1046–1052.

Utne, K.R., Huse, G., Ottersen, G., Holst, J.C., Zabavnikov, V., Jacobsen, J.A., Óskarsson, G.J., Nøttestad, L., 2012b. Horizontal distribution and overlap of planktivorous fish stocks in the Norwegian Sea during summers 1995-2006. Marine Biological Research, 8, 420-441.

Vilhjálmsson, H., 1994. The Icelandic capelin stock. Rit Fiskideildar 13, 1-281. Vilhjálmsson, H., 2002. Capelin (Mallotus villosus) in the Iceland-East Greeland-Jan Mayen

ecosystem. ICES Journal of Marine Science 59, 870-883.


23

Ware, D.M., Lambert, T.C., 1985. Early life history of Atlantic mackerel (Scomber scombrus) in the southern Gulf of St. Lawrence. Canadian Journal of Fisheries and Aquatic Science. 42, 577-592.

Wiborg, K.F., 1955. Zooplankton in relation to hydrography in the Norwegian Sea. Report on Norwegian Fishery and Marine Investigations 1-66.

Wilson, P.C., Bruce, R.A., Stevenson, W.H., Dixwell Chase, E., 1965. Cruise report, MV Delaware cruise 65-3; March 30 - April 23, 1965. Exploratory Fishing and gear research base, Bureau of Commercial Fisheries, Fish and Wildlife Service, US Dept. of Interior, Gloucester, MA.

Yabe H, Ueyanagi, S., Kikawa, S., Watanabe, H., 1959. Study on the life-history of the swordfish, Xiphias gladius Linnaeus.Rep. Nankai Reg. Fish. Res. Lab. 10, 107-150.

Yaragina, N.A., Marshall, C.T. 2000. Trophic influences on interannual and seasonal variation in the liver condition index of Northeast Arctic cod (Gadus morhua). ICES Journal of Marine Science, 57, 42–55.


24

2 Stomach Contents of Herring, Blue Whiting, Mackerel, Albacore and Bluefin Tuna in the North Atlantic Earth System Science Data (ESSD) Paper – EuroBasin Special Issue John K. Pinnegar1, Nicholas Goñi3, Verena M. Trenkel2, Haritz Arrizabalaga3, Webjørn Melle4, James Keating5, Guðmundur Óskarsson6. 1Centre for Environment, Fisheries & Aquaculture Science (Cefas), Pakefield Road, Lowestoft, Suffolk, NR33 0HT, UK. [Tel. +44 (0) 1502 524229, Fax. +44 (0) 1502 513865, e-mail, [email protected]] 2 Institut français de recherche pour l'exploitation de la mer (IFREMER), BP 21105, 44311 Nantes, France 3

AZTI Fundazioa, Herrera kaia e Portualdea z/g, 20110 Pasaia, Gipuzkoa, Spain. 4Institute of Marine Research (IMR), P.O. Box 1870 Nordnes, 5817 Bergen, Norway. 5

Commercial Fisheries Research Group, Galway Mayo Institute of Technology (GMIT), Galway, Ireland 6 Marine Research Institute (MRI), Skulagata 4, 121 Reykjavik, Iceland 1. Introduction In recent years considerable emphasis has been dedicated towards finding 'ecosystem-based' approaches to fisheries management and multispecies models are seen as crucial for addressing this new agenda. However, there are very few long-term predator-prey datasets, available within a European context for parameterising such models and characterising food-web interactions in general. Marine food-webs have become a major focus of EU research and maritime policy in recent years. The 2008 European Marine Strategy Framework Directive (2008/56/EC) included a requirement for Member States to work to achieve ‘Good Environmental Status’ (GES) by 2015. This is defined by eleven qualitative descriptors, one of which (descriptor 4) explicitly focuses on “Food Webs”. In addition recent documents on reform of the EU ‘Common Fisheries Policy’ (e.g. COM(2011) 417), have stated that “Fisheries management must ... follow the ecosystem and precautionary approach” and this has been interpreted as needing to take account of interactions between species. Therefore there is growing demand for information on ‘who eats whom’ in marine systems, in order to deduce how changes in one part of the ecosystem might have consequences elsewhere.

Maximum sustainable yield (MSY) is the optimal catch that can be taken from a fish stock year after year without endangering its capacity to regenerate for the future. EU Member States made a commitment at the World Summit on Sustainable Development in 2002 to work towards MSY for all fish stocks by 2015. However, modelling studies (such as Mackinson et al. 2009) have demonstrated that it is highly unlikely that all stocks can be maintained at precautionary MSY reference points simultaneously. In reality, the very high yields predicted at low fishing pressure by single-species stock assessment models would be eroded by predation pressure. Consequently, ICES have stated that “Stomach data are of vital importance” and that it intends to gradually

mailto:[email protected]


25

transition to providing multispecies advice on fisheries for some ecosystems in the near future (ICES 2013). A number of coordinated stomach databases do exist in Europe but these typically encompass only a limited selection of species or cover a very narrow time period. One of the more extensive datasets is the ICES ‘Year of the Stomach’ database for the North Sea, which provides information on 35 species, although detailed information is only available for 9, primarily based on stomachs collected during sampling campaigns in 1981 and 1991. A similar coordinated ICES dataset exists for cod in the Baltic Sea and has been documented in ICES (1997). In the Barents Sea, a combined database exists for Norway and Russia (Dolgov et al., 2007), but in all cases these sources offer limited information on pelagic fish with a wide geographic distribution. The EU EURO-BASIN project aims to understand and predict the dynamics of plankton and pelagic fish species in the North Atlantic, and to assess the impacts of climate variability. This project, funded under the EU FP7 programme, pays particular attention to herring (Clupea harengus), mackerel (Scomber scombrus) and blue whiting (Micromesistius poutassou), which are the most abundant and widespread planktivorous fish species in the region, but also bluefin tuna (Thunnus thynnus) and albacore (Thunnus alalunga), top predator species that inhabit the whole North Atlantic basin and carry out large transatlantic migrations. In the present paper we provide details of newly digitised information on the diet of these five species, firstly from the UK DAPSTOM database (mackerel, herring and blue whiting), that incorporates information from recent research cruises by the Institute of Marine Research – IMR (Norway), Marine Research Institute - MRI-HAFRO (Iceland), Institut français de recherche pour l'exploitation de la mer - IFREMER (France), - Marine Institute, (Ireland), as well as historic data from the Centre for Environment, Fisheries & Aquaculture Science (Cefas) in the UK. We also include data from AZTI-Tecnalia (Spain) on bluefin tuna and albacore stomach contents. Datasets derived under the EURO-BASIN project have been submitted to the PANGAEA open-access data portal (www.pangaea.de). 2. Data and Methods The DAPSTOM database The DAPSTOM database has been in existence for 8 years, having been created in response to a ‘data-rescue’ call from the EU ‘Network of Excellence’ project EUROCEANS. The most recent version of the DAPSTOM dataset (Version 4.5, collated in September 2013) includes 207,907 records derived from 396 distinct research cruises, spanning the period 1837- 2012. The database contains information from 237,617 individual predator stomachs and 184 species. As such, this represents one of the largest and most diverse compilations of food-web data anywhere in the world. A key component of the DAPSTOM programme has been development of an online data portal (www.cefas.defra.gov.uk/fisheries-information/fish-stomach-records.aspx) through which datasets are made freely available to the wider scientific community. As the DAPSTOM initiative has progressed, a relational-database structure has evolved in Microsoft Access that can accommodate most formats of stomach content information (see Hyslop 1980 for a review), including data collected at the level of individual fish, pooled samples of multiple fish

http://www.cefas.defra.gov.uk/fisheries-information/fish-stomach-records.aspx


26

stomachs, frequency of occurrence data as well as fully gravimetric information (prey weights or volumes). As a bare minimum, in Version 4.5 of the DAPSTOM database, information on the predator species, geographic area and the number of stomachs was required for a dataset to be included. Information on predator length (or size range) was also usually available. Central to the relational-database structure is the “DAPSTOM” data table (see figure 1). This includes much of the ‘raw’ information about both the predator and prey. The “DAPSTOM” data table includes 23 information fields (see figure 1), and a full definition of each field is provided by Pinnegar and Platts (2011). Important linking variables include “Haul ID” a unique identifier which is also replicated in the main ‘Hauls’ table, typically comprising the cruise name and station number.”Pred” is the predator species, indicated by a 3 digit code (and linked to the ‘Predator’ look-up table); “Prey” is the prey type, as written in the original paper source, this is used as a linking variable to the ‘Prey’ look-up table; “PRED ID” is a unique identifier for the individual predator animal (or group of animals). This is needed because there is sometimes more than one prey item within a particular stomach. The “MIN NUM” field is important for database extractions and outputs. Where the number of prey items is given in the original source, then this is reproduced in this field. Where no prey number is given, then a minimum of ‘1’ is assumed. The “HAULS” table contains all information about the geographic location from which the sample was derived. In most cases this includes details of the ship name, dates and times, latitudes, longitudes, depths, gear type, ICES area and any additional information. As a bare minimum each ‘haul’ must be assigned to predefined “Sea” (e.g. North Sea, Irish Sea, W Ireland, Celtic Sea, Channel, Biscay etc.) and ICES “Division” – a spatial sub-unit used by the International Council for the Exploration of the Seas. The ‘provenance look-up’ table (see below) is linked to the “HAULS” table via the “Cruise Name” (see figure 1). A new innovation within version 4.5 of the DAPSTOM database is the inclusion of a ‘PROVENANCE’ look-up table (see figure 1). The purpose of this is to record the original source of the data that has been digitised. The PROVENANCE table is linked to the ‘HAULS’ table and acknowledges the person (at Cefas or elsewhere) who made the information available. Two additional look-up tables have been created to help standardise the taxonomic information that is available to users (see below). The “PREDATOR” look-up table expands on the 3 digit codes in the “DAPSTOM” table and gives the predator’s latin name, common name (in English), 10 digit NODC code and TSN identifier. The “PREY” look-up table aims to reduce the enormous number of potential prey names and descriptions to a manageable number of standardized names that can then be used for analyses and collation. It corrects historic taxonomy to modern counterparts, and allows aggregation by broad prey groups (e.g. euphausids, amphipods, copepods, teleosts etc.).


27

‘HAULS’

•Year•Cruise Name•Station

•Haul ID•Gear Type•Date•NewDate•ICES Rectangle

•ICES Sub-area•ICES Division•ICES Roundfish area•Sea•Shot Lat Degrees

•Shot Lat Minutes•Shot Long Degrees•Shot Long Minutes•Shot East/West•Haul Lat Degrees

•Haul Lat Minutes•Haul Long Degrees•Haul Long Minutes•Haul East/West•Shot Time•Haul Time•Depth•Comments

‘PROVENENCE’

•Cruise Name•Batch•Source Type•Data Input•Original Source

‘DAPSTOM’

•Haul ID•Pred•Pred Length

•Pred Sex•Pred Maturity•Pred Weight•Pooled (y/n)•Size Category•Mean Length•Pred ID•Number of Stomachs•Number Empty•Fullness•Total Stomach Weight•Total Stomach Volume•Prey•Prey Number•Prey Length•Digestion Stage•Ind Prey Weight•Ind Prey Volume•Minimum Number

‘PREDATOR’

•Pred•Scientific Name•Common Name•NODC Code•TSN Code

‘PREY’

•Prey•MAFF•Valid Name•Scientific Name•Group•Type•NODC Code

Figure 1. Relational structure of the DAPSTOM 4.5 database, including a list of the fields included in each of the five linked tables. Pelagic fish data submitted to PANGAEA under the EU project Euro-Basin The DAPSTOM database was explicitly mentioned in the ‘Description of Work’ for Euro-Basin, as ‘an open-access repository to accommodate datasets generated during the project’. Throughout 2013, Euro-Basin partners submitted datasets and these were reformatted into the required relational tables (see figure 1). Datasets made available as part of Euro-Basin can be summarised as follows: Table1. Number of records for pelagic fish species submitted to the DAPSTOM database as part of Euro-Basin. Number of individual stomachs included in parentheses.

Dataset Herring Blue whiting Mackerel Albacore Bluefin tuna

IFREMER (France) 0 (0) 133 (117) 0 (0) 0 (0) 0 (0)

IMR (Norway) 1291 (538) 354 (139) 1772 (635) 0 (0) 0 (0)

MRI (Iceland) 1610 (823) 274 (158) 3226 (1486) 0 (0) 0 (0)

GMIT (Ireland) 0 (0) 139 (109) 0 (0) 0 (0) 0 (0)

Cefas (2010-2011) 1101 (961) 467 (366) 6 (3) 0 (0) 0 (0)

Cefas - Historical 4506 (25424) 216 (237) 5614 (5299) 1 (1) 10 (3)

Total 8508 (27746) 1583 (1126) 10618 (7423) 1 (1) 10 (3)

From table 1 it is clear that the vast majority of the blue whiting data within the DAPSTOM database, were explicitly collected for the purposes of Euro-Basin (1367 records out of 1583), whereas this was not true for either herring or mackerel. It is also apparent that for blue whiting and mackerel the number of database records exceeded the number of stomachs examined,


28

confirming that the data were largely non-pooled records from individual stomachs whereas this was not true for herring, where 8,508 database records were derived from 27,746 stomachs. The primary explanation for this disparity is the digitisation of ‘pooled’ herring datasets from a historical report by Hardy (1924), but also ‘pooled’ data from Brook & Calderwood (1886) and Scott (1924). Table 2 shows the number of records and samples by geographic area, including all larval and juvenile fish. From this table it is apparent that herring, blue whiting and mackerel have been sampled over a huge geographic area, from the Bay of Biscay (~43°N), to the high Arctic (~73°N) and from Greenland in the West (~29°W) to the Lofoten islands in the East (~9°E). By contrast the very limited number of records for Albacore and Bluefin tuna in the DAPSTOM database were derived from the English Channel and North Sea (although see the section on albacore and blue-fin tuna below). Table 2. Number of records for pelagic fish species submitted to the DAPSTOM database as part of Euro-Basin, by geographic region.

Sea Herring Blue whiting Mackerel Albacore Bluefin tuna

Biscay 0 157 896 0 0

Celtic Sea 66 506 2804 0 0

Channel 577 35 718 1 1

E Greenland 605 70 1050 0 0

Iceland 680 105 1356 0 0

Irish Sea 1294 183 29 0 0

North Atlantic 0 18 0 0 0

North Sea 2954 19 1106 0 9

Norwegian Sea 1616 435 2447 0 0

West Ireland 0 55 129 0 0

West Scotland 716 0 83 0 0

The earliest data included in the Euro-Basin dataset is a single record of albacore stranded on the Channel coast of England in August 1864, whereas the most recent data comes from a single bluefin tuna stranded at Ventnor, Isle of Wight in August 2012. The pelagic fish dataset includes information on the feeding preferences of fish larvae (0.1 to 13 cm in length), as well as juvenile and adult fish. Specifically, the feeding habits of larval/juvenile herring and mackerel from Plymouth Sound, the Clyde and the North Sea by Lebour (1921, 1924), Marshall et al. (1937, 1939) and Last (1980) respectively, were digitised. 3. Results & Discussion The DAPSTOM dataset has now seen wide usage among ICES Working Groups as well as in a number of theoretical ecology papers (e.g. Rochet et al. 2011; Rossberg et al. 2011; Brose et al. 2006). On the whole, researchers have used the online portal to look at the diet composition of their favoured predator species – however there has also been some interest in making use of historical datasets to look at long-term changes in fish diets at particular localities (Le Quesne &


29

Pinnegar, 2012). A major limitation of the DAPSTOM dataset is that it comprises a mixture of ‘pooled’ information and data collected from individual fish. Sometimes only information on the number of stomachs containing a particular prey item out of the total number of stomach examined was available (i.e. ‘frequency of occurrence’), rather than the actual number of a particular prey item. Hence in any data extraction, such as the one illustrated in table 1, outputs should be viewed as providing information on the ‘minimum number’ of prey items consumed. This would have little impact in predator species that consume large prey items (e.g. fish feeders), and in most of the newer datasets assembled under EuroBasin, but it could mean that in certain older datasets the total number of prey items in plankton-eating species such as mackerel, herring and blue whiting would be grossly underestimated. An example of where this might be the case is the historical dataset on mackerel stomachs off the Cornish coast, digitised from Bullen 1908, as well as the herring datasets digitised from Marshall et al. (1937, 1939) and from Hardy (1924). A further limitation of the DAPSTOM database is the sporadic availability of information on prey weights. In many of the constituent datasets no gravimetric information was provided. A result of this is that it can be difficult to judge the importance of a particular prey item to the overall nourishment of the predator, since a mackerel for example, may draw significantly more nourishment from eating a single fish in comparison with 1000+ copepods. To remedy this situation, a long-term aspiration of the DAPSTOM project is to develop an updated ‘PREY’ table that includes average prey weights, and perhaps energy density for each standardised prey type so that numbers consumed can be converted to total weights – but this feature is not yet available. Several authors have suggested that the preferred prey of blue whiting are euphausiids and hyperiid amphipods, although the relative importance of each of these varies depending on season and locality (e.g. Prokopchuk & Sentyabov 2006; Langøy et al. 2012 ). The EuroBasin dataset (table 1 and 2) shows similar variability in diet composition depending on sampling location (figure 2), with euphausids dominating in terms of numerical dominance in Iceland, the Bay of Biscay and the Irish Sea, but hyperiid amphipods dominating in the Norwegian Sea, eastern Greenland and the Celtic Sea. Copepods (mainly Calanus finmarchicus) were an additional important prey item in the Norwegian Sea and shrimps (in particular Pasiphaea sivado) were commonly observed in stomachs from the Irish Sea. Adult blue whiting migrate in the springtime, to the Porcupine and Rockall areas west of Ireland. During this season they feed very infrequently. Post larval mackerel feed on a variety of zooplankton and small fish. Published sources suggest that the main zooplankton prey organisms in the North Sea are copepods (mainly Calanus finmarchicus), and euphausiids (mainly Meganyctiphanes norvegica), while fish prey include larval sandeel, herring and sprat (Mehl and Westgård 1983). In the Norwegian Sea published sources suggest that euphausiids, copepods, pteropod molluscs (Limacina retroversa), amphipods, appendicularia and capelin are the main dietary items (Langoy, et al 2012; Prokopchuk & Sentyabov 2006). The EuroBasin dataset confirms these broad patterns (figure 3), and in every geographic region for which data were available (Table 2), with the exception of the Bay of Biscay, copepods dominated in terms of numerical abundance, especially in Iceland (91%). Mysids and hyperiid amphipods contributed a significant additional proportion to the diet of mackerel in the North Sea (16%), hyperiids and euphausids contributed a significant additional proportion in east Greenland (32%) and phytoplankton, teleosts and chaetognaths contributed a significant proportion in the Celtic Sea (27%). In the Bay of Biscay, 67% of the diet composition (by number) was suggested to comprise mackerel eggs (denoted as ‘teleosts’ in figure 3), although the vast majority of these data


30

come from a single research cruise in March 1986 and from a very limited number of haul stations. Mackerel stop feeding almost completely during winter. For herring, many detailed diet composition studies of have been published, starting with Hardy (1924). In the Norwegian Sea diet has been shown to vary depending on availability of food and geographic location (Prokopchuk & Sentyabov 2006; Langøy et al. 2012). C. finmarchicus is an important prey in summer (about 77% by weight). But in certain years appendicularians (Oikopleura spp.), amphipods (mainly Parathemisto abissorum), and euphausiids are important. In the EuroBasin dataset, copepods dominated herring diets in terms of numerical abundance in the Norwegian Sea, Iceland, North Sea and West of Scotland (69, 85, 66, 74% respectively, figure 4). Hyperiid amphipods contributed a significant additional proportion in the Norwegian Sea (24%), appendicularians contributed a significant additional proportion in the North Sea (15%) and barnacle cypris larvae a significant additional proportion in the west of Scotland (21%). In eastern Greenland, euphasiids were the dominant prey item (63%), followed by copepods (16%) and hyperiid amphipods (15%). In the Irish Sea euphausiids comprised 49% of the diet and fish eggs (denoted as ‘teleosts’ in figure 4, but mostly plaice Pleuronectes platessa) contributed a further 31%. However, these Irish Sea research cruises (in February 2009, 2010 and 2011) were deliberately timed to quantify the seasonal predation mortality imparted by pelagic fish on plaice eggs and larvae. Albacore diet composition in the northeast Atlantic has been reported as being dominated by small, mesopelagic fish, e.g. Maurolicus muelleri and Scomberesox saurus, but also euphausiids and hyperiid amphipods (Pusineri et al 2005; Goñi et al. 2011). Bluefin tuna diet in the northeast Atlantic has been reported by Logan, et al. (2011). In the Bay of Biscay, euphausids (Meganyctiphanes norvegica) and anchovy (Engraulis encrasicolus) made up 39% prey weight, with relative consumption of each reflecting annual changes in prey abundance. These same data, as well as more recent information have been submitted to PANGAEA as part of the EuroBasin project. Bluefin tuna used to be distributed in the North Sea, from where it disappeared in the 1960s, it is thought that they fed primarily on herring and on mackerel in this region (Tiews 1978). Within EuroBasin, the datasets described in this paper will be used to calculate overall consumption exerted by pelagic fish on particular mesozooplankton taxa in the northeast Atlantic. Daily ration estimates will be generated by applying the methodology of Pennington (1985). These values will then be used to quantify ‘top down’ predation pressure in the different regional seas. The longer-term aspiration by making these datasets available (through PANGAEA and DAPSTOM) is this they will facilitate the construction of more realistic ecosystem or fisheries models that can then be used to provide holistic advice as now required by the EU and international conventions.

http://icesjms.oxfordjournals.org/search?author1=C.+Pusineri&sortspec=date&submit=Submit


31

VIIa (Irish Sea)Amphipod

Euphausiid

Copepod

UI Crustacean

Teleost

Pteropod

Shrimp

Other

XIVa (Greenland)Va (Iceland) IIa (Norwegian Sea)

VIII (Biscay) VIIe-j (Celtic Sea) VIIa (Irish Sea)

Figure 2. Diet composition of blue whiting Micromesistius poutassou in different parts of the northeast Atlantic, based on data collated as part of the EuroBasin project and submitted to DAPSTOM/PANGAEA. Proportions are based on the number of individual prey items.

IV (North Sea)Copepod

Pteropod

Amphipod

Euphausid

Phytoplankton

Teleost

Appendicularia

Mollusc

Chaetognath

Mysid

Cladocera

Crustacean

Jellyfish

Other

Va (Iceland) IIa (Norwegian Sea) XIV (Greenland)

VIII (Biscay) VII (Celtic Sea) IV (North Sea)

Figure 3. Diet composition of mackerel Scomber scombrus in different parts of the northeast Atlantic, based on data collated as part of the EuroBasin project and submitted to DAPSTOM/PANGAEA. Proportions are based on the number of individual prey items.


32

XIV (Greenland) Copepod

Appendicularia

Amphipod

Chaetognath

Pteropod

Euphausid

Teleost

UI Crustacean

Phytoplankton

Barnacle

Cladocera

Mollusc

Other

Va (Iceland) IIa (Norwegian Sea) XIV (Greenland)

VIIa (Irish Sea) VIa (West Scotland) IV (North Sea)

Figure 4. Diet composition of herring Clupea harengus in different parts of the northeast Atlantic, based on data collated as part of the EuroBasin project and submitted to DAPSTOM/PANGAEA. Proportions are based on the number of individual prey items.

Acknowledgements Initial development of the DAPSTOM data portal was supported through a ‘data rescue’ grant from the EU Network of Excellence programme ‘EurOceans’. Further iterations were supported by research grants from the UK Department for Environment, Food & Rural Affairs (specifically contracts MF1202 and MF1109). Ongoing developments have been supported by work-package 5 of the EuroBasin project. References Brook, G. and Calderwood, W.L. Report on the food of herring. In: Appendix F, VI - Preliminary

reports on the food of fishes. Fourth Annual Report of the Fishery Board for Scotland, being for the year 1885, 102-128, 1886.

Brose, U., Jonsson, T., Berlow, E.L., Warren, P., Banasek-Richter, C., Bersier, L.F., Blanchard, J.L., Brey, T., Carpenter, S.R., Blandenier, M.F., Cushing, L., Dawah, H.A., Dell, T., Edwards, F., Harper-Smith, S., Jacob, U., Ledger, M.E., Martinez, N.D., Memmott, J., Mintenbeck, K., Pinnegar, J.K., Rall, B.C., Rayner, T.S., Reuman, D.C., Ruess, L., Ulrich, W., Williams, R.J., Woodward, G. and Cohen, J.E. Consumer–resource body-size relationships in natural food webs. Ecology, 87, 2411–2417, 2006.

Bullen, G.E. Plankton studies in relation to the western mackerel fishery. J. Mar. Biol. Assoc. U.K., 8, 269-302, 1908.

Dolgov, A. V., Yaragina, N. A., Orlova, E. L., Bogstad, B., Johannesen, E., and Mehl, S. 20th anniversary of the PINRO–IMR cooperation in the investigations of feeding in the Barents Sea—results and perspectives. In: Proceedings of the 12th Norwegian–Russian Symposium, Tromsø, 21–22 August 2007, pp. 44–78. IMR/PINRO Report Series, 5/2007, 2007.

Goñi, N., Logan, J., Arrizabalaga, H., Jarry, M. and Lutcavage, M. Variability of albacore (Thunnus alalunga) diet in the Northeast Atlantic and Mediterranean Sea. Mar. Biol., 158: 1057-1073,


33

2011. Hardy, A.C. The herring in relation to its animate Environment. Part I. The food and feeding habits

of the herring with special reference to the east coast of England. MAFF Fishery Investigations II, 7, 1-39, 1924.

Hyslop, E.J. Stomach contents analysis – a review of methods and their application. J. Fish Biol., 17, 411-429, 1980.

ICES. Report of the study group on multispecies model implementation in the Baltic. International Council for the Exploration of the Sea. ICES CM 1997/J:2, 1997.

ICES. Multispecies considerations for the North Sea stocks. ICES Advice 2013, Book 6, section 6.3.1, June 2013. International Council for the Exploration of the Sea, Copenhagen, Denmark, 2013.

Langøy, H., Nøttestad, L., Skaret, G., Broms, C. and Fernö, A. Overlap in distribution and diets of Atlantic mackerel (Scomber scombrus), Norwegian spring-spawning herring (Clupea harengus) and blue whiting (Micromesistius poutassou) in the Norwegian Sea during late summer, Mar. Biol. Res., 8, 442-460, 2012.

Last, J.M. The food of twenty species of fish larvae in the west-central North Sea. MAFF Fisheries Research Technical Report 60, pp 44, 1980.

Lebour, M.V. The food of young clupeoids. J. Mar. Biol. Assoc. U.K., 12, 458-467, 1921. Lebour, M.V. The food of young herrings. J. Mar. Biol. Assoc. U.K., 13: 325-330, 1924. Le Quesne, W.J.F. and Pinnegar, J.K. The potential impacts of ocean acidification: scaling from

physiology to fisheries. Fish. Fisher. 13, 333-344, 2012. Logan, J.M., Rodríguez-Marín, E., Goñi, N., Barreiro, S., Arrizabalaga, H., Golet, W., and Lutcavage,

M. Diet of young Atlantic bluefin tuna (Thunnus thynnus) in eastern and western Atlantic foraging grounds. Mar. Biol., 158, 73-85, 2011.

Mackinson, S., Deas, B., Beveridge, D and Casey, J. Mixed-fishery or ecosystem conundrum? Multi-species considerations inform thinking on long-term management of North Sea demersal stocks. Can. J. Fish. Aquat. Sci. 66, 1107–1129, 2009.

Marshall, S.R., Nicholls, A.G. and Orr, A.P. On the growth and feeding of the larval and post-larval stages of the Clyde herring. J. Mar. Biol. Assoc. U.K., 22, 245-267, 1937.

Marshall, S.R., Nicholls, A.G., and Orr, A.P. On the growth and feeding of young herring in the Clyde. J. Mar. Biol. Assoc. U.K., 23, 427-455, 1939.

Mehl, S. and Westgård, T. The diet and consumption of mackerel in the North Sea. International Council for the Exploration of the Sea. ICES CM 1983/H: 34, 1983.

Pennington, M. Estimating the average food consumption by fish in the field from stomach contents data. Dana, 5, 81–86, 1985.

Pinnegar, J.K. and Platts, M. DAPSTOM - An Integrated Database & Portal for Fish Stomach Records - Version 3.6. Centre for Environment, Fisheries & Aquaculture Science (Cefas), Lowestoft, UK. July 2011, pp 35, 2011.

Pusineri, C., Vasseur, Y., Hassani, S., Meynier, L., Spitz, J., and Ridoux, V. Food and feeding ecology of juvenile albacore, Thunnus alalunga, off the Bay of Biscay: a case study. ICES J. Mar. Sci., 62, 116-122, 2005.

Prokopchuk, I. and Sentyabov, E. Diets of herring, mackerel, and blue whiting in the Norwegian Sea in relation to Calanus finmarchicus distribution and temperature conditions. ICES J. Mar. Sci., 63, 117-127, 2006.

Rochet, M.-J., Collie, J.S., Jennings, S. and Hall, S.J. Does selective fishing conserve community biodiversity? Predictions from a length-based multispecies model. Can. J. Fish. Aquat. Sci. 68, 469-486, 2011.


34

Rossberg, A.G., Farnsworth, K.D., Satoh, K. and Pinnegar, J.K. Universal power-law diet partitioning by marine fish and squid, with surprising stability-diversity implications. P. Roy. Soc., B-Biol. Sci., 278, 1617-1625, 2011.

Scott, A. Food of the Irish Sea Herring in 1923. In: Proceedings and Transactions of the Liverpool Biological Society. Vol XXXVIII. pp 115-119, 1924.

Tiews, K. On the disappearance of bluefin tuna in the North Sea and its ecological implications for herring and mackerel. Rap. Procés., 172, 301–309, 1978.


35

3 Analysis of stomach samples of small pelagic fish in the northern North Sea Laura Meskendahl Introduction The purpose of the present investigation was to quantify the consumption by herring and associated pelagic fish stocks such as sprat and mackerel in the northern North Sea. The interactions between planktonic prey and fish predators affect early life stage survival, e.g. mackerel preying on blue whiting juveniles. Hence, these interactions will be quantified and the competition between the three pelagic species will be assessed. This activity requires distribution and abundance estimates of predators, diet compositions, and the development of consumption rate models as well as prey production rates. In the present report, results from a brief study are presented, which were originally thought to contribute to the understanding of top-down effects of small pelagic fishes in adjacent waters to the Atlantic Ocean. The main objective was to quantify the stomach content of sprat and herring. Due to the low sample size of mackerel no further comparative analysis with stomach contents of this species was possible. Hence, the main goals were 1) possible competition between sprat and herring and 2) investigating the diet composition and stomach fullness of sprat from the northern North Sea in comparison with fish caught in the German Bight. Materials and Methods Sampling Samples were taken during the WH352 cruise in March 2012. A bottom trawl net GOV (Chalut á Grande Ouverture Verticale), was employed to obtain the groundfish abundance. Also small pelagic fish such as herring (Clupea harengus L.) and sprat (Sprattus sprattus L.) were caught and used as study fish for the present investigation. Fish were frozen on board and stomachs were dissected in the laboratory. Only samples taken in the northern most transects were analyzed here (Fig. 1). Stomach analysis For each fish total length, standard length and the weight (0.01 g) before and after stomach removal were noted. For each stomach, full and empty weight was measured and the content preserved in 70% ethanol. From max 30 fish per length class (total length; cm) stomachs were dissected. For the qualitative analysis of the stomach samples, each probe was analyzed under a binocular microscope. An estimate of the volumetric percentage of highly digested items was recorded as these could not be quantified by numbers. It was assumed that the countable items were also representative for the digested portion of the stomach contents.Only clearly identifiable prey was counted as numbers and the digestion stage was determined (Table 1). If possible, length measurements from up to ten individuals of a prey group were conducted on digital images. If the stomach sample clearly contained only one prey group, all intact prey was counted and additionally


36

eyes (or pairs of eyes) were counted if possible (e.g. from Euphausiacea; see Fig. X). Under the assumption that all pairs of eyes belong to the non identifiable rest, the total number of prey can be estimated by the sum of eye pairs and the counted prey individuals. The qualitative analysis was conducted only for some stomachs in order to get an idea of the preferred prey types/sizes. Most of the analyzed stomach contents was, however, only weighted and preserved in ethanol. Table 1: Indices for the qualitative analysis of stomach contents from sprat and herring

Digestion Stage 0= intact prey 1= prey with damages, but only slightly digested 2= prey with damages, not clearly identifiable verdGrad45 = not identifiable, highly digested

Prey Numbers Counted numbers of prey in the stomach

% prey Estimated percent of prey from digested material

Figure 1: Cruise track of WH 352. ICES roundfish areas (numbered in grey). Red circles indicate the selected stations from which stomach contents were analyzed. For sprat two additional stations were included in the analysis (St. 394 and 399, circles on the left map). Results and Discussion Sprat


37

A total number of 495 sprat stomachs were weighted from fish between 9 and 14 cm total length. At Station 285 all fish were feeding, whereas on station 289 (very close to the coast) the majority of fishes had empty stomachs and only max 20% were feeding. Samples from other stations always contained some individuals with empty stomachs, but the majority of fish were feeding (Table 2). However, with reference to the results from all analyzed stations, there was no general pattern explaining empty or full stomachs. Sampling on station 289 was conducted very early in the morning (06:26) and close to the coast (see Fig. 1), but it is unclear if the empty stomachs were an effect of low light level in the morning or low food availability due to other environmental conditions. Table 2. Overview of analyzed stomach samples of sprat (Sprattus sprattus L.) from the northern North Sea and two additional stations further south (St. 394 and 399, see Fig. 1).

Station LK fish feeding non-feeding fish total fish % feeding fish

285 9 7 0 7 100

10 30 0 30 100

11 28 0 28 100

12 11 0 11 100

13 8 0 8 100

14 8 0 8 100

289 10 0 20 20 0

11 6 24 30 20

12 2 19 21 10

292 9 30 0 30 100

10 30 0 30 100

11 28 0 28 100

12 11 1 12 92

320 9 15 5 20 75

10 10 9 19 53

11 5 17 22 23

12 2 13 15 13

394 7 20 0 20 100

8 20 0 20 100

9 17 3 20 85

10 17 4 21 81

399 7 8 12 20 40

8 8 12 20 40

9 17 3 20 85

10 7 4 11 64

Note: LK = length class (cm), total fish = total number of fish analyzed, fish feeding = fish with measurable stomach content, non-feeding fish = fish with empty stomachs The relative stomach content (% of body wet weight) decreased with increasing length class (Fig. 2), indicating that younger and smaller fish fed faster, presumably due to higher metabolic demands


38

compared to larger conspecifics. Prey items found in sprat stomachs were mainly small cirrepedia nauplia, copepods or zoea larvae. (Fig. 3; Table 3). Only four stomachs were analyzed in detail.

Figure 2. Stomach content (% of body wet weight) of feeding sprat per length class (cm) from all sampling stations (given in each panel). Blue points represent the mean, bars indicate the 25 and 75-%-percentile, respectively.

Table 3. Numbers of prey found in sprat stomachs in the northern North Sea.

Station Fish-Nr. PREY Taxon PREY group PREY stage Stage of digestion prey [%] prey total

Figure 3. Examples of prey items found in stomachs of sprat (Sprattus sprattus L.) from the northern North Sea sampled in March 2012.


39

292 3 Crustacea Cirripedia Nauplius 0-1 50 751

292 3 Crustacea Copepoda adults; copepodites 0-1

8

292 3 unknown unknown unknown 0-1

8

292 3 n.i. verdGrad45 50

292 8 Crustacea Cirripedia Nauplius 0-1 50 746


292 6 Crustacea Cirripedia Nauplius 0-1

487

292 6 Crustacea Copepoda adults; copepodites 0-1

160

292 6 Crustacea Decapoda Zoea-Larvae 0-1

40

292 6 Crustacea Decapoda unknown 2

3


285 5 Crustacea Copepoda adults; copepodites 0-1 0.5

285 5 Crustacea unknown

1 0.5 80


Note: n.i. = non identifiable; verdGrad45 = see Table 1; prey total = total number of prey calculated from the counted prey and the volumetric percentage (prey %) of counted and non identifiable prey.

Figure 4. Stomach content (g) of feeding sprat per length class (cm) from all sampling stations (given in each panel). Blue points represent the mean, bars indicate the 25 and 75-%-percentile, respectively.


40

The comparison of sprat stomach content data collected in March 2012 (Fig. 4) and those collected during the GLOBEC-Germany program in spring (February – April) 2003 and 2004 (Fig. 5) revealed much higher mean stomach contents (g) in the 2012 samples. Within the samples from the GLOBEC program some high values occurred (up to 0.5 g), but the mean values were always below 0.1. In contrast, the stomachs sampled in 2012 have mean values around 0.2 g and some high values of ~ 0.8 g have been found. This might be an effect of much higher prey densities in the northern areas of the North Sea and close to the Dogger Bank (stations 394 and 399) than in the German Bight, which was the main sampling area during the GLOBEC program or can be affected by the sampling time (10 year difference between both samples). However, no further analysis was conducted to explain the differences in mean stomach contents or to test the significance of the difference in stomach contents between the two sampling periods. Clupea harengus

Due to the low number of fish sampled per station (app. 4 kg per station, regardless of the number or size of the individuals), only two stations were analyzed exemplary. Overall 115 herring stomachs from fish between 13 and 31 cm total length were weighted (Table 4). The size overlap between sprat and herring on the same station was low, so that no direct comparison of diet composition between these two species was possible. On station 289 almost no fish were feeding, whereas on station 292 all fish smaller than 23 cm in length were feeding, but some larger individuals did not contain food in their stomachs (see Table 4; Fig. 6 & 7).

Figure 5. Stomach content (g) of feeding sprat per length class (cm) from all sampling stations in the German Bight in February to April 2003 and 2004 from the GLOBEC-Germany project (unpublished data). Blue points represent the mean, bars indicate the 25 and 75-%-percentile, respectively.


41

Table 4. Overview of analyzed stomach samples of herring (Clupea harengus L.) from the northern North Sea.

Station LK fish feeding non-feeding fish total fish % feeding fish

289 13 0 1 1 0

14 0 17 17 0

15 2 23 25 8

16 0 9 9 0

17 0 4 4 0

19 0 2 2 0

292 14 1 0 1 100

15 3 0 3 100

16 3 0 3 100

17 11 0 11 100

18 11 0 11 100

19 5 0 5 100

20 3 0 3 100

21 4 0 4 100

22 1 0 1 100

23 3 0 3 100

24 1 1 2 50

25 1 1 2 50

26 0 2 2 0

27 0 4 4 0

31 0 2 2 0

Note: LK = length class (cm), total fish = total number of fish analyzed, fish feeding = fish with measurable stomach content, non-feeding fish = fish with empty stomachs


42

Figure 6. Stomach content (% of body ww) of feeding herring per length class (cm) from both sampling stations (given in each panel). Blue points represent the mean, bars indicate the 25 and 75-%-percentile, respectively.

Figure 7. Stomach content (% of body ww) of feeding herring per length class (cm) from both sampling stations (given in each panel). Blue points represent the mean, bars indicate the 25 and 75-%-percentile, respectively.


43

4 Variability of albacore and bluefin tunas top-down trophic impacts in the Northeast Atlantic Nicolas Goñi, Haritz Arrizabalaga – AZTI-Tecnalia, Herrera kaia portualdea z/g, 20110 Pasaia (Spain) Introduction Albacore and bluefin tuna are highly migratory species present in the whole temperate North Atlantic (Santiago, 2004 ; Fromentin and Powers, 2005), and exploited by surface and subsurface fisheries during summer months in the Bay of Biscay and surrounding waters. Both species are important pelagic predators targeting a large number of crustaceans, fish and cephalopods species (Aloncle and Delaporte, 1974, Logan et al., 2011), and having a wide geographical distribution in the Eastern North Atlantic in summer months. Their predation impact may therefore be important and affect a high number of prey communities in this region. This impact may undergo important fluctuations at an interannual level, depending on the recruitment fluctuations of prey species, and on variations in spatial distributions of both predators and of prey. The present study focuses more particularly on three important small pelagic fish species in their juvenile stages: Atlantic saury, blue whiting and anchovy. These relatively high caloric species have a potential high importance in albacore diet in terms of energy input, thus the variations of their relative proportions in albacore diet may have strong implications in albacore feeding success. Moreover, two of them (blue whiting and anchovy) also support important commercial fisheries, thus better information on their mortality by predation at juvenile stages is relevant to understand recruitment variability. Materials and methods Stomachs of 1448 albacore and of 834 bluefin tunas were collected during 2004 to 2007 and during 2009 to 2012. The first period corresponds to a situation of collapse of the stock of anchovy in the Bay of Biscay. The second period corresponds to a situation of recovery of the anchovy stock and to an anomaly in albacore summer distribution, with a very scarce presence in the Bay of Biscay and a higher local abundance in Southwestern Irish waters. Albacore stomachs were sampled in five different geographic areas in 2004-2007 and in twelve areas (broadest sampling done) in 2010. The sampling areas comprise both shelf-break and oceanic areas, and the albacore originated from the three main fisheries operating in the region, i.e. baitboat, trolling and pelagic trawling fisheries. Each stomach was stored frozen, then thawed before weighting. Its contents were identified at the lowest possible taxonomic level, using keys based on morphological characteristics by Ibañez Artica et al. (1989), and the online database www.fishbase.org (Froese and Pauly, 2010) in the case of fish prey. Crustaceans were identified using the manual by Todd et al. (1996), which comprises morphological descriptions of crustacean species. Cephalopods were identified by the morphological characteristics of their beaks, according to the handbook by Clarke (1986). Each taxonomic group was weighed and its corresponding number of individuals counted when possible. The weight proportion of each prey group in each stomach was calculated, as well as its mean weight proportion by sampling area and year. Results and Discussion

Albacore diet displays a high plasticity with an important spatial variability, both latitudinally and in terms of oceanic vs shelf-break waters. The most ubiquitous prey species (present in most sampled areas) was krill Meganyctiphanes norvegica in both periods. Among small pelagic fish species, Atlantic saury, blue whiting and anchovy were the major prey in the shelf-break areas of the Bay of Biscay and Celtic Sea.

Atlantic saury consumption showed a latitudinal variability. This prey represents a larger proportion


44

of albacore diet in the northern sampling areas than within the Bay of Biscay. This corroborates previous observations by Aloncle and Delaporte (1974), who related the presence of Atlantic saury to relatively low SST values mainly found out of the Bay of Biscay. This result suggests a potential higher predation impact on Atlantic saury when the summer distribution of albacore shifts westwards as in 2009-2011. Saury was also present in bluefin diet in 2011 and – to a lesser extent – in 2012, in which it seemed to partly replace anchovy as a caloric prey.

Blue whiting consumption appears related to the shelf-break of the Bay of Biscay and Celtic Sea but does not appear to vary significantly along with latitude. However its consumption displays very important interannual variations, with a particularly low presence in albacore diet in 2010 and 2011 compared to 2004-2007. Taking into account the decrease in blue whiting biomass in recent years (Payne et al., 2012), this decrease in blue whiting consumption by albacore is more likely to be simply related to a lower availability of the prey rather than to a shift in feeding preferences. This suggests that blue whiting predation by albacore is globally proportional to blue whiting abundance, and that albacore do not tend to select blue whiting among other available prey species. A complementary interpretation would be a lower availability of blue whiting (closely related to shelf-break areas) to albacore when its summer distribution was displaced westwards in 2009-2011. This is corroborated by the observed important proportion of blue whiting in the diet of bluefin tuna, which were sampled mostly in shelf-break locations.

Anchovy consumption displays an important seasonal and latitudinal variability, being higher in the late summer and autumn in the southern Bay of Biscay. The same seasonal variability appears in bluefin tuna diet. This spatial and seasonal variability of anchovy consumption by both predators is related to anchovy life-cycle and to the ecology of juvenile stages (spawning on the continental shelf during spring, juveniles getting out of the shelf to oceanic waters from early August onwards). Within the Bay of Biscay, we observed a broader spatial distribution of anchovy in albacore diet during 2009-2011 than during 2004-2007. This broader distribution was probably related to the recovery of anchovy population, after a period of collapse between 2005 and 2008. The combined variability of the spatial extension of juvenile anchovies and of albacore distribution in summer months suggests a distinct spatial match/mismatch and predation impact each year.

In terms of daily consumption of anchovy, it varied (according to the year and season) from 6.7 to 83.1 individuals (max. 188) per day and predator for bluefin tunas. This consumption was lower for albacore, ranging between 4.8 and 16.5 (max. 103). Summary and main findings Stomachs of 1448 albacore and of 834 bluefin tunas were collected during 2004 to 2007 and during 2009 to 2012 (130 bluefin stomachs were also sampled in 2013). The first period corresponds to a situation of collapse of the stock of anchovy in the Bay of Biscay. The second period corresponds to a situation of recovery of the anchovy stock and to an anomaly in albacore summer distribution, with a very scarce presence in the Bay of Biscay and a higher local abundance in Southwestern Irish waters. Concents of all stomachs were analyzed. Albacore diet displays a high plasticity and an important spatial variability both latitudinally and in terms of oceanic vs shelf-break waters. Bluefin diet displays an important pattern of seasonal variability. The present study focuses more particularly on Atlantic saury, blue whiting and anchovy, which – among small pelagic fishes – are the main prey of albacore and bluefin tunas. Consumption of the three species displays an important interannual variability, probably related to interannual recruitment variations. Atlantic saury consumption by albacore also shows a latitudinal variability, appearing more important in the northern part of albacore summer distribution. Blue whiting consumption appears related to the shelf-break of the Bay of Biscay and Celtic Sea but not significantly to latitude. Anchovy consumption displays an important seasonal variability for both albacore and bluefin tuna, with a higher consumption rate in the second part of the summer and in early autumn. Anchovy consumption by albacore also displays a latitudinal variability, being higher in the southern Bay of Biscay. In spite of the recovery of anchovy population, its daily consumption by both tuna species still displays a relatively important inter-annual variability.


45

The stomach content data of albacore and bluefin tunas will be uploaded to a Pangaea database (part of the WP1 tasks). These data will help quantifying the role of key species and functional groups in biomass and carbon transfers within the marine food web, and evaluating the relative importance and interactions of top down versus bottom up controls on community structure and carbon sequestration. (WP4). They will also help characterizing the distribution of food web types and key species in the Bay of Biscay oceanic and shelf-break waters (WP3) in terms of abundance, biomass and size spectra. Finally, they will provide hopefully useful information for the basin-scale modeling done within WP6 tasks. References Aloncle H, Delaporte F (1974) Données nouvelles sur le germon Atlantique Thunnus alalunga

Bonnaterre 1788 dans le Nord-Est Atlantique. 1ère Partie – Rythmes alimentaires et circadiens. Revue des Travaux de l’Institut des Pêches Maritimes 37 (4) : 475-572.

Clarke MR (1986) A handbook for the identification of cephalopod beaks. Clarendon Press, Oxford, UK, 273 pp.

Froese R, Pauly D (Editors) (2010). FishBase. World Wide Web electronic publication. www.fishbase.org, version (05/2010).

Fromentin JM, Powers JE (2005) Atlantic bluefin tuna: population dynamics, ecology, fisheries and management. Fish Fish 6:281–306

Ibañez Artica M, Menendez de la Hoz M, Matallanas J, Ramos A, Sanchez F, San Millan MD (1989). Euskal Herriko arrainak. ISBN: 978-84-7728-128-3, Editorial Kriselu S.A., Donostia (Spain), 112 pp.

Logan J, Rodríguez-Marín E, Goñi N, Barreiro S, Arrizabalaga H, Golet W, Lutcavage M (2011). Diet of young Atlantic bluefin tuna (Thunnus thynnus) in eastern and western Atlantic forage grounds. Marine Biology 158(1): 73-85.

Santiago J (2004) Dinámica de la población de atún blanco (Thunnus alalunga Bonaterre 1788) del Atlántico Norte. PhD Thesis, Euskal Herriko Unibertsitatea, Bilbao, 320 pp.

Todd CD, Laverack MS, Boxshall G (1996) Coastal Marine Zooplankton: a practical manual for students. Cambridge, University Press. 2nd edition, 106 pp.

http://www.fishbase.org/


46

5 Spatially explicit estimates of prey consumption of the North Atlantic albacore Tuna (Thunnus alalunga) P. Lehodey, I. Senina, A.C. Dragon

CLS, Marine Ecosystem Department, 31520 Ramonville, France

The model SEAPODYM (Lehodey et al 2008) was applied to the north Atlantic albacore tuna to model the spatial population dynamics of the species over the historical fishing period 1960-2008 (Dragon et al. in prep.; Lehodey et al 2013; Lehodey et a., in prep.) Using Maximum Likelihood Estimation approach (Senina et al 2008), the model parameters were optimized using fishing data, allowing to provide realistic distribution of fish density by cohort and life stage at a resolution of 2°x month. The model predicted spatio-temporal variability of fish density under the combined effect of fishing activity and climate variability. Once optimized with fishing data, it can be run also without fishing effort to measure the impact of fishing. The dynamics of the species is linked to the definition of a feeding habitat relying on the simulation of six functional groups of micronekton (Lehodey et al 2010) distributed over three vertical layers (Fig. 1). This model is currently revised (WP6) to improve its parameterization based on acoustic data (Lehodey et al., in prep). The feeding habitat accounts for the accessibility by the fish to the micronekton biomass in each layer according to the fish preference (e.g. temperature) or limitation (eg oxygen concentration). Thus using a daily food ration estimate (ie an average value of 5% of body weight), the model predicted the consumption of albacore tuna in time and space and by life stage (even cohort) for each group of prey (Fig. 1). Therefore it is possible to investigate the change over time and oceanic region, and compare as well the change with a simulation without fishing impact. The results showed how the diet is predicted to change while fish are growing and can access deeper colder layers. This is particularly clear in the tropical and subtropical region characterized by very high biomass of migrant and non migrant mesopelagic micronekton. In the temperate region diet of young and adult fish are roughly similar, however with less non migrant bathypelagic organisms for young fish. For a same region the composition in the diet by % of functional group does not show significant change over time (Fig. 1), nor when comparing simulation results with and without fishing (not shown). It is worth noting however that the functional groups are defined on vertical behaviour of organisms and cannot account for change in the type of species or group of species that compose these groups. In other terms, the model makes the assumption of a large flexibility of tuna in their diet, which seems a reasonable hypothesis given the opportunistic feeding behavior observed for all tuna species. Finally the total consumption of micronekton groups can be computed over time (Fig. 1) to be compared to the micronekton production. Our preliminary results suggest that tuna consume a very small percentage of the micronekton production, ranging between 0.3 and 1.6% according to the group and area. However this result is very sensitive to the energy transfert coefficients used in the micronekton model and thus will need to be revised after the optimal parameterization using assimilation of acoustic data will be achieved (link with WP6). This study will be continued in the coming months and extended with the projection under climate change scenarios. It would be


47

useful also to confront these simulations results with actual data as those compiled from recent stomach analyses (Goni et al. this report) but also possibly from older studies and from other regions, especially the tropical area.

a

b

c: 1961-1970

fishing

d: 1991-2000 fishing

e

f

Figure 1. Identification of micronekton functional groups on acoustic echogram (a) and (b)SEAPODYM (conceptual) model (Lehodey et al. in prep.), and predicted average composition (%) of diet by young immature (left pie) and mature adult (right pie) according to these prey groups in four oceanic regions of the Atlantic Ocean and two decades (c and d). Total (young + adult) consumption by prey group over time are shown for areas 1 (e) and 4 (f). References Dragon AC, Senina I, Lehodey P., Arrizabalaga H. (in prep). Modeling spatial populaiton dynamics of


48

North Atlantic Albacore tuna under the influence of both fishing and climate variability. Lehodey P, Senina I, Dragon A-C, Arrizabalaga H (in prep). Spatially explicit estimates of stock size,

structure and biomass of North Atlantic albacore Tuna (Thunnus alalunga). To be submitted to special issue ESSD EURO-BASIN.

Lehodey P, Senina I, Dragon A-C, Arrizabalaga H. (2013). Modeling activities conducted under EURO-BASIN research project to develop SEAPODYM to the North Atlantic albacore tuna (Thunnus alalunga) ICCAT, Document SCRS/2013/ 125: 26 pp.

Lehodey P., Conchon A, Senina I, Domokos R, Calmettes B, Jouanno J, Hernandez O, Kloser R. (in prep). Optimization and evaluation of a micronekton model with acoustic data.

Lehodey P., Murtugudde R., Senina I. (2010). Bridging the gap from ocean models to population dynamics of large marine predators: a model of mid-trophic functional groups. Progress in Oceanography, 84: 69–84

Lehodey P., Senina I., Murtugudde R. (2008). A Spatial Ecosystem And Populations Dynamics Model (SEAPODYM) - Modelling of tuna and tuna-like populations. Progress in Oceanography, 78: 304-318.

Senina I., Sibert J., Lehodey P. (2008). Parameter estimation for basin-scale ecosystem-linked population models of large pelagic predators: application to skipjack tuna. Progress in Oceanography, 78: 319-335.


49

6 Trophic impact of top predators migrations in exploited environments

Patrizio Mariani1*

, Ken H Andersen1, Brian R MacKenzie

2

1 Centre for Ocean Life, and 2 Centre for Macroecology, Evolution and Climate, National Institute for Aquatic

Resources, Technical University of Denmark, Charlottenlund, Denmark

* corresponding author: [email protected]

1. Introduction

Trophic relationships between living organisms form the backbone of ecosystem functioning and

biodiversity (Paine 1966, Worm and Duffy 2003). Several factors can affect the magnitude and

importance of these trophic interactions driving non-linear processes and complex dynamics in the

food webs (Levin 1998).

When trophic links are strong, perturbations in one component of the network can have cascading

effects in all other components (Pace et al 1999). For example changes in the abundance of

individuals in one trophic level can elicit direct and indirect changes on other trophic levels, a

process that known as trophic cascade (Paine 1980, Pace et al 1999, Polis et al 2000). Evidences of

trophic cascades have been reported both in terrestrial and in aquatic ecosystems (Pace et al 1999,

Shurin et al 2002, Schmitz et al 2004) supporting the hypothesis of a widespread process in

ecosystem dynamics. However ecologists have often debated about how ubiquitous trophic cascades

are in ecosystems (Polis 1994, Polis et al 2000). This is also because several compensatory

mechanisms can dampen or eliminate trophic cascades (Pace et al 1999, Cury et al 2003, Andersen

& Pedersen, 2010, Heath et al 2014).

In marine food webs the high degree of connectivity and the ubiquitous presence of omnivory and

ontogenic diet shift can prevent or dampen trophic cascades (Baum and Worm 2009, Andersen and

Pedersen, 2010). For large groups such as pelagic fish where all individuals span several trophic

levels from larval to the adult stages, such compensatory mechanisms can be quite important (Reid

et al. 2000) limiting the possibilities of trophic cascades in this group.

However, behavioral traits are central in the functioning of pelagic ecosystems and can drive trophic

cascades in marine communities (Werner and Peacor, 2003). For example, long distance migration

can produce large perturbations in local food web by increasing in a relative narrow period predation

pressure on some of its component (Polis et al 1996). Casini et al (2012) showed that when high

abundances of Baltic cod Gadus morhua migrated into unoccupied habitats in the Baltic Sea, their

predatory impact induced four-level trophic cascades in the forage fish, zooplankton, and

phytoplankton communities of the region.

Moreover, fishing activities can also be an important factor in changing the dynamics of marine food

webs and can then significantly regulate trophic cascades (Andersen and Pedersen, 2010). High

fishing pressures have been suggest to potentially trigger regime shifts in large marine ecosystems

(Jackson et al 2001, Daskalov et al 2007) and are likely responsible for the recent changes in the fish

community structure (Jennings et al 1999). Unperturbed ecosystems of some decades ago would

have been most likely characterized by a vast number of large predators and negligible fishing

pressures. Hence, major differences between today’s and pristine ecosystems can be expected



50

(Steneck and Sala, 2005) posing questions about how present food webs have been shaped by the

recent trophic downgrading (Estes et al. 2011).

Atlantic bluefin tuna is one of those large apex predators that in some marine ecosystem have been

largely removed by fishing. Here, we analyze the trophic impact of such removal, by simulating the

effects in the northern North Sea food-web when exposed to the historical seasonal migration of

Atlantic Bluefin tuna. This species used to migrate into the North Sea region for many years in the

early-mid 1900s but stopped in the mid-1960s and has been rare or non-existent during most years

ever since (Mather et al. 1995; MacKenzie and Myers 2007; Fromentin 2009). Bluefin tuna would

migrate mainly for the Mediterranean Sea spawning ground and would remain in the region for 2-4

months before migrating southward in the autumn. While present in the North Sea, their

consumption of prey (assumed primarily to be mackerel and herring) has been estimated to be at his

maximum up to 500 kt (Tiews 1978).

Here we investigate whether such predation could have impacts on trophic levels farther down the

food web and evaluate the effects a renewed migration of tuna into the North Sea. We use a size- and

trait-based model for the fish community that is able to account for the changes in trophic levels

during ontogeny and resolve fishing mortality on larger sizes (Andersen and Pedersen 2010). The

model is used to reconstruct the North Sea fish community size spectra under different scenarios of

tuna migrations and fishing pressure and allows estimating trophic cascades induced by migrations

of the top predator.

Methods

Model formulation

We use a previously published size-spectrum model to represent the base-line fish community

(Andersen and Pedersen, 2010; see appendix for concise description including equations and

parameters). The model is a physiologically structured model (de Roos and Persson, 2001) based on

a description of the energy budget of individuals. All rates and processes are parameterized using the

size of individuals and the asymptotic size of species (Hartvig et al 2011, Appendix A). The result of

the model is the distribution of individual abundances as function of size and asymptotic size :

) and its variation over time, . The central process in the model is predation of small

individuals by large individuals. The food obtained from predation fuels growth and reproduction.

Reproduction is limited by a stock-recruitment relationship to ensure coexistence of a continuum of

asymptotic sizes.

The impact of Atlantic bluefin tuna (henceforth tuna) on the resident fish community is represented

in the model as an extra mortality inflicted in the size range defined by the prey size preference

of tuna where is the weight of prey. The total consumption of tuna can be calculated

as:

(1)

where is the community spectrum which is the integral over all asymptotic sizes:

.

The equation for can be rewritten to isolate the mortality:

(2)


51

Fishing mortality is imposed as a function of individual size and asymptotic size. All asymptotic

sizes are fished with a sigmoid relationship modulated with a common fishing mortality, which has

been calibrated to emulate the fishing pattern in the North Sea (Pope et al 2006).

We present the results in terms of four scenarios: unperturbed ecosystems with no fishing or tuna

migrations, i.e., F = 0, Stuna = 0 ( ); unexploited ecosystem with no fishing but tuna migration, i.e.,

F = 0, Stuna =150 kton ( ); exploited ecosystem with high fishing but no tunas, F = 0.7, Stuna = 0

( ); exploited ecosystem with tunas, i.e., F = 0.7, Stuna = 150 kton ( ). Moreover, we also

analyse the trophic impact of fishing and migrations for a large range of fishing mortalities and tuna

consumptions. We assume that the prey range of tuna is constrained within a minimum value

( gr) and some large size ( Kg), i.e., .

We us an index to measure the perturbation on the fish community size-structure produced by

migration of tuna. This index is defined as the integral of the oscillations in the size-spectra

relative to the case of no tuna feeding:

(3)

where indicates the specific fishing scenario considered (unexploited or exploited at different

levels). This index is positive defined and measure the integral change in size structure as driven by

tuna migration and feeding for all sizes smaller than the maximum tuna’s prey size ( ).

Feeding consumption of tuna in the northern North Sea

Estimating the prey consumption by tuna requires knowledge of how many bluefin tuna were

present, what size they were and their daily ration. No abundance estimates are yet available for this

time period. However by combining commercial catch data (ICCAT 2012) and estimates of fishing

mortality rate (Fromentin and Restrepo 2009) for the most important fishery in the region (i. .e., the

Norwegian fishing fleet), it is possible to estimate biomass. We assumed as an initial approximation

that the fishing mortality rate represents the removal rate of biomass:

Ftuna = catch/B (4)

where B is the total biomass. Fishing mortality rates were estimated to be F = 0.3, F= 0.2 and F = 0.1

per year, for the years 1950s, 1960s and 1970s respectively (Fromentin and Restrepo 2009). As most

of the individuals captured in this fishery were adults (Tiews 1978, Fromentin and Restrepo 2009),

the biomass would correspond only to a spawning stock biomass (i.e., no juvenile biomass) in this

region. Indeed the reported weights in the catches (Tiews 1978) were mainly in the range 150 – 400

Kg. We used two reference tuna weights, w1 = 200 Kg and w2 = 300 Kg, to derive the number of fish

in the area (N) and the average daily consumptions of the population (Stuna):

Stuna = 0.5 k (w10.8

N1+ w20.8

N2) (5)

where Stuna is expressed in Kg, k = 0.123 (Innes et al, 1987, Overholtz 2005) and N1,2 are the number

of fish in class w1,2, i.e., N1,2 = B/w1,2.

The consumption equation was originally derived from a meta-anlyses of consumption rates in


52

marine mammals (Innes et al, 1987) and has been previously used to estimate consumption rates in

bluefin tunas (Overholtz 2005). Based on the relation above we obtain an average individual

consumption rate of 4% body weight (BW). The daily ration of bluefin tuna in the North Sea during

the 1950s-60s has been previously assumed to be 3-6% BW/day (Tiews 1978) and 1-4.7% BW/day

(Overholtz 2005), which are then consistent with the estimate obtained above.

The spawner biomass estimates we derive are based only on Norwegian catch data, and exclude the

catches of other fishing nations (e. g., Denmark, Germany, Sweden), which account for 15-20% of

the total catches (Tiews 1978). Hence, the biomass estimates and associated population-level

consumption rates of prey will be most likely underestimates. Nevertheless, we provide sensitivity

analyses of the model results to changes in Stuna.


53

Results

Feeding of tuna in the northern North Sea

Norwegian landings of tuna in the region varied between 15 kton in 1952 to 0 after the 1987. This

corresponds to an estimated maximum biomass of about 50 kton.

We can then estimate for each year the total consumption by tuna assuming a residence period of

100 days (Figure 1). This consumption is estimated to be a between 100 -150 kton with maximum

value of 200 kton in 1952. Moreover following the catches data the consumption decreases to around

25 kton after 1963 and then to 0 in most recent periods. Those values compare well with previous

most conservative estimates of tuna consumption in the area (Tiews 1978). Hence, large oscillations

in the trophic impact of tuna in the area can be expected.

Figure 1. Total biomass consumed by bluefin tuna migrating in the northern North Sea between 1950 and 2009. Our

values (blue line) are compared with a series of estimates from Tiews (1978) (black dashed lines). Catches data of the

Norwegian fleet have been transformed to total biomass and then into daily consumption. The total consumed biomass is

then calculated assuming 100 days of feeding in the area and is used to define the model parameter Stuna.

Community response to top predator migration and fishing

At the equilibrium and under unperturbed conditions ( , F = 0, Stuna = 0) the simulated ecological

community in the northern North Sea distributes according to a size spectra with a slope of -1.47

(Figure 2) and a spawning stock biomass between 50 kton and 400 kton for the smaller and larger

asymptotic size classes, respectively.


54

weight (g)

Figure 2. Number size spectra of 10 asymptotic size classes in an unperturbed ecosystem (black lines) and community

spectrum (red line) and the zooplankton resource spectrum (green line). To represent the northern North Sea, the total

biomass of the ecosystem is set to half of the values estimated for the entire North Sea, i.e., we use a biomass of 5 x103

kton.

Considering a pristine northern North Sea characterized by presence of migrating tuna populations

but absence of significant fishing mortality (F = 0) we can compare this pristine ecosystem ( )

with the unperturbed scenario above (Figure 3a). The assumed consumption of the migratory

population of tunas can substantially reduce the abundance of the larger prey in the feeding range (w

> 1Kg), but will have a positive impact to the smaller prey items (0.1 Kg < w < 0.3 Kg). This is

mainly because smaller sizes will experience a reduction of both predation pressure and competition

from the larger sizes. The trophic impact from tuna will cascade down to smaller individuals in the

size spectra model with dampening effects the closer we move into the region dominated by resource

biomass (w < 0.5 g).

The simulated pristine ecosystem is substantially different from the simulated fish community when

a more recent scenario with fishing pressure is considered ( , Figure 3a). In case of a fishing

mortality at F = 0.7 and no tuna migrations (Stuna = 0) a large increase of fish biomass (1.2 times) in

the range 0.1 Kg < w < 1 Kg is simulated in the model. As for the previous case the increase of

biomass in certain range of sizes has cascading effects in the community size spectra with decreasing

biomass for the individuals at about w = 10-2

Kg, and an increase for those at about 10-4

Kg , while

negligible effects are predicted for variation of the smaller fish species in the model (Figure 3a). This

scenario is robust to additional migrations of top predators ( , Figure 3a), which has the only

effect to move slightly the peaks of these oscillations towards smaller sizes.

At equilibrium the highest prey biomass in the feeding range of tuna (0.1 Kg < w < 2.5 Kg) is

obtained in the exploited case with no tuna ( ) while the opposite conditions (no fishing but tuna

) provides the lowest. The net effect of tuna feeding on the fish community can be directly

measured by comparing the cases with and without tuna migration at different fishing mortality

(Figure 3b). As shown before, the introduction of tuna feeding on the fish community as a

detrimental effect for the larger tuna’s prey, while the smaller sizes in the prey range benefit of some

decrease in competition and then increase their biomass. However, the impact is weaker when the

community is also exposed to fishing (Figure 3b, black line). Indeed, the perturbation on the


55

community size spectra introduced by tuna migration is dampened the further we move towards

smaller sizes.

Figure 3. Relative change in community structure for the scenario analysed at different individual weights. In (a)

unexploited ecosystem with tuna migration ( , green) exploited ecosystem with no tuna ( , black) and exploited

ecosystem with tuna migration ( , red) are compared to the unperturbed ecosystem scenario ( , dashed line). In (b)

the impact of tuna migration on the ecosystem is assed using the ratio (black) and (blue). The tuna,

when present, is feeding in the range 0.1 < w <2.5 Kg (grey area) with a consumption of Stuna = 150 kton, while fishing

mortality on the ecosystem is F = 0.7. The integral index to assess impact on the size structure is graphically shown as

the area under relative change curve (blue shaded are in b).

Model sensitivity

Given a perturbation scenario with both fishing and tuna migration ( ), we can calculate the

change in the tuna prey ( ) as the difference in biomass between and the

unperturbed case, i.e., .

In case of F = 0 the introduction of tuna in the ecosystem can generate a lower equilibrium biomass

in their feeding range (Figure 4a). A sharp transition to low value of biomass (-500 kton) is present

in the model for tuna consumption around Stuna = 400 kton. At this feeding level the prey biomass

reaches a critical minimum that cannot sustain the recruitment of larger species. But, when the

number of the larger fishes shrinks, the tuna prey will benefit of some reduced predation and will

then slightly increase in biomass; just enough to keep a continuous flow of biomass on the size-

spectra. Hence, at this critical level, the size spectra-model will have an oscillatory behaviour in the

tuna prey range. The average size-spectra in the critical region can have up to 10% relative standard

error, which is a measure of the range of variability in the model. However, those oscillations are for

a very low biomass and further increase in tuna migration will not affect the solution.

Introducing fishing will generally have positive effects at the size range of the tuna prey. Indeed at

moderate level of fishing ( )) and no tuna migration, there is a maximum increase of 500

kton of tuna prey biomass produced. Nonetheless, further increases in fishing mortality will also

affect smaller sizes including the tuna prey size. The same conclusions can be drawn when both tuna

migration and fishing are present in the model, resulting in a general non-linear relation between the

two processes (Fig 4a). Interestingly, at low level of fishing mortality (F=0.5) up to 600 kton of tuna

feeding can be present in the model without significantly affecting the tuna prey biomass compared


56

to an unperturbed scenario.

The direct impact of tuna migration in the northern North Sea ecosystem can be evaluated, at

different values of tuna consumption, using the perturbation index (Fig 4b). Since is a measure

relative to the absence of tuna migration (Eq. 3), the index start from at all level of fishing

mortality and can be generally described as the area below a perturbation curve (Fig 3b). Tuna can

significantly affect the size structure of the model, , at high level of consumption or at low

level of fishing mortality (Fig 4b). Indeed if F is high the effect of tuna migration on the size-

structure is negligible. While at low levels, , the larger the tuna consumption the greater is

the effect on the food-web; up to the critical transition described above, where the collapse of the

tuna prey range will largely impact the size-structure and any further increase in Stuna will not have

effects.

Figure 4. Map of the (a) tuna prey equilibrium biomass (in kton) and (b) size-structure index (Eq. 3) at different values

of tuna consumption (Stuna) and fishing mortality (F). The tuna prey biomass is calculated in the prey interval

, with gr and Kg, while the is defined as the changes in size structure for

.

Discussions

General discussion

Our results suggest that pristine environments characterized by low fishing and significant presence

of large carnivores could have been quite different from the marine ecosystems we have today. In

our model the pristine ecosystems have a total biomass of small pelagics that is directly controlled

by predation of large top predators, such as bluefin tuna in the northern North Sea. The reduced

biomass in this group can have significant effects on other components of the food web through a

trophic cascading effects mechanism.

When fishing is introduced in the system the large predators are greatly reduced in biomass up to the

point that they are entirely removed them from the community. Thus, at some intermediate value of

fishing there are beneficial effects on small pelagics biomass, since removal of predators can reduce

the feeding pressure on the group. Nonetheless at high fishing mortalities also the biomass of the

small pelagic communities will be reduced.

The results suggest that in the northern North Sea there is now more prey biomass for tuna than there


57

was when they were used to migrate in the area. This could support the hypothesis that with a correct

management strategy, the population of tuna can migrate back in the region of the North Sea, re-

establishing the migration path that has been currently lost.

Nonetheless, as previously shown (De Luca et al. 2014), the sudden interruption of migration

towards certain feeding areas is also explained by a lost of group collective memory generated by

breaking the social transmission of the information. The removal from a group of those individuals

that have some information and preference for a given feeding areas can produce rapid changes in

the migratory behaviour of the school and can make the group moving towards other destination

sites or not being able to migrate at all (De Luca et al. 2014). Hence, given that food resources are

available for tuna, the species can rebuild the migratory path towards the northern North Sea when

some vagrant individual is allowed to rediscover this historical feeding area.

There are ample evidences that migratory species can explore the environment and discover new

areas for feeding (Alerstam et al 2003). Recently, migrations of mackerel and bluefin tuna species

have been observed in western part of Greenland, an area that was not previously used by either

species (MacKenzie et al 2014). Those changes might have been driven by improved conditions of

the habitats in northern areas (e.g., increase in temperature, increase in food) and can now have

cascading effects on the local fish community. We suggest that the fish community in Greenland can

be exposed to significant changes in biomass and size-structure as the seasonal migration of these

top predators is established as a route for a large fraction of the migratory population. This

hypothesis can be likely tested in the future.

On the contrary, the re-introduction of tuna in the northern North Sea would have negligible effects

on both size-structure and pelagic biomass, because the fishing pressure in the area is much greater

than the potential impact of tuna feeding.

Although it is recognized that predation in marine ecosystems is an important factor explaining food

webs dynamics and community structures (Verity and Smetacek 1996, Heithaus et al 2008, Baum

and Worm 2009, Mariani et al 2013), effects of top predators on the marine food web are more

complex because of non linear interactions and vulnerability to fishing (Cury et al 2003, Ferretti et al

2012).

Our results suggest that significant trophic cascades can be driven by seasonal migrations of large

top predators. However, those effects are only present in ecosystems that are subject to no- or low-

fishing mortality. Indeed, we demonstate that re-introduction of Atlantic bluefin tuna in the present

state North Sea ecosystem will not have any further impact on the community structure, since the

fish community is already highly exposed to fishing pressure.

Acknowledgements

The research leading to these results has received support from the EU-FP7 project EURO-BASIN

(grant agreement no. 264933).

References

Alerstam, T., Hedenström, A., & Åkesson, S. (2003). Long‐distance migration: evolution and determinants. Oikos,

103(2), 247-260.

Andersen, K. H., & Pedersen, M. (2010). Damped trophic cascades driven by fishing in model marine

ecosystems. Proceedings of the Royal Society B: Biological Sciences, 277(1682), 795-802.


58

Baum, J. K., & Worm, B. (2009). Cascading top‐down effects of changing oceanic predator abundances. Journal of

Animal Ecology, 78(4), 699-714.

Casini, M., Blenckner, T., Möllmann, C., Gårdmark, A., Lindegren, M., Llope, M., ... & Stenseth, N. C. (2012). Predator

transitory spillover induces trophic cascades in ecological sinks. Proceedings of the National Academy of

Sciences, 109(21), 8185-8189.

Cury, P., Shannon, L., & Shin, Y. J. (2003). The functioning of marine ecosystems: a fisheries perspective. Responsible

fisheries in the marine ecosystem, 103-123.

De Luca, G., Mariani, P., MacKenzie, B. R., & Marsili, M. (2014). Fishing out collective memory of migratory schools.

Journal of The Royal Society Interface, 11(95), 20140043.

de Roos, A.M., Persson, L., 2001. Physiologically structured models - from versatile technique to ecological theory.

Oikos 94, 51–71.

ICCAT 2012. Report of the 2012 Atlantic Bluefin Tuna Stock Assessment Session. Doc. No. SCI-033 / 2012, p. 124.

Innes, S., Lavigne, D. M., Earle, W. M., & Kovacs, K. M. (1987). Feeding rates of seals and whales. The Journal of

Animal Ecology, 115-130.

Fromentin, J. M., & Restrepo, V. (2009). A year-class curve analysis to estimate mortality of Atlantic bluefin tuna caught

by the Norwegian fishery from 1956 to 1979. Collective Volume of Scientific Papers, 64(2), 480-490.

Ferretti, F., Osio, G. C., Jenkins, C. J., Rosenberg, A. A., & Lotze, H. K. (2013). Long-term change in a meso-predator

community in response to prolonged and heterogeneous human impact. Scientific reports, 3.

Hartvig, M., Andersen, K. H., & Beyer, J. E. (2011). Food web framework for size-structured populations. Journal of

Theoretical Biology, 272(1), 113-122.

Heath, M.l R., D. C. Speirs, and J. H. Steele. "Understanding patterns and processes in models of trophic cascades."

Ecology letters 17.1 (2014): 101-114.

Heithaus, M. R., Frid, A., Wirsing, A. J., & Worm, B. (2008). Predicting ecological consequences of marine top predator

declines. Trends in Ecology & Evolution, 23(4), 202-210.

Jennings, S., Greenstreet, S., & Reynolds, J. (1999). Structural change in an exploited fish community: a consequence of

differential fishing effects on species with contrasting life histories. Journal of Animal Ecology, 68(3), 617-627.

Levin, S. A. (1998). Ecosystems and the biosphere as complex adaptive systems. Ecosystems, 1(5), 431-436.

Mariani, P., & Visser, A. W. (2010). Optimization and emergence in marine ecosystem models. Progress in

Oceanography, 84(1), 89-92.

MacKenzie, B. R., Payne, M. R., Boje, J., Høyer, J. L., & Siegstad, H. (2014). A cascade of warming impacts brings

bluefin tuna to Greenland waters. Global change biology.

Overholtz, W. J. (2006). Estimates of consumption of Atlantic herring (Clupea harengus) by bluefin tuna (Thunnus

thynnus) during 1970–2002: an approach incorporating uncertainty. J. Northwest Atl. Fish. Sci, 36, 55-63.

Pace, M. L., Cole, J. J., Carpenter, S. R., & Kitchell, J. F. (1999). Trophic cascades revealed in diverse

ecosystems. Trends in ecology & evolution,14(12), 483-488.

Paine, R. T. (1966). Food web complexity and species diversity. American Naturalist, 65-75.

Paine, R. T. (1980). Food webs: linkage, interaction strength and community infrastructure. Journal of Animal

Ecology, 49(3), 667-685.

Polis, G. A. (1994). Food webs, trophic cascades and community structure. Australian Journal of Ecology, 19(2), 121-


59

136.

Polis, G. A., Holt, R. D., Menge, B. A., & Winemiller, K. O. (1996). Time, space, and life history: influences on food

webs. In Food Webs (pp. 435-460). Springer US.

Polis, G. A., Sears, A. L., Huxel, G. R., Strong, D. R., & Maron, J. (2000). When is a trophic cascade a trophic

cascade?. Trends in Ecology & Evolution,15(11), 473-475.

Pope, J. G., Rice, J. C., Daan, N., Jennings, S. & Gislason, H. 2006 Modelling an exploited marine fish community with

15 parameters—results from a simple size-based model. ICES J. Mar. Sci. 63, 1029–1044.

Schmitz, O. J., Krivan, V., & Ovadia, O. (2004). Trophic cascades: the primacy of trait‐mediated indirect

interactions. Ecology Letters, 7(2), 153-163.

Shurin, J. B., Borer, E. T., Seabloom, E. W., Anderson, K., Blanchette, C. A., Broitman, B., ... & Halpern, B. S. (2002).

A cross‐ecosystem comparison of the strength of trophic cascades. Ecology letters, 5(6), 785-791.

Steneck, R., & Sala, E. (2005). Large marine carnivores: trophic cascades and top-down controls in coastal ecosystems

past and present (pp. 110-37). University of Maine Sea Grant College Program.

Verity, P. G., & Smetacek, V. (1996). Organism life cycles, predation, and the structure of marine pelagic

ecosystems. Marine ecology progress series. Oldendorf, 130(1), 277-293.

Werner, E. E., & Peacor, S. D. (2003). A review of trait-mediated indirect interactions in ecological communities.

Ecology, 84(5), 1083-1100.

Worm, B., & Duffy, J. E. (2003). Biodiversity, productivity and stability in real food webs. Trends in Ecology &

Evolution, 18(12), 628-632.


60

Appendix A

Encounter and consumption

Prey size selection

M1

Volumetric search rate

M2

Encountered food

M3

Maximum consumption rate

M4

Feeding level

M5

Growth and reproduction

Maturation function

M6

Somatic growth

M7

Egg production

M8

Recruitment

Population egg production

M9

Recruitment

M10

Mortality

Background mortality

M11

Predation mortality

M12

Fishing selectivity

M13

Fishing mortality

M14

Resource spectrum

Growth rate

M15

Carrying capacity

M16

7 Consumption of Calanus finmarchicus by planktivorous fish in the


61

Norwegian Sea Kjell Rong Utne1,2, Solfrid Sætre Hjøllo1, Geir Huse1 and Morten Skogen1 INTRODUCTION The Norwegian Sea ecosystem The Norwegian Sea is a part of the Nordic Seas, located north of the North Sea, east of the Icelandic Sea, southeast of the Greenland Sea and west of the Barents Sea (Fig. 1). The average depth of 1800 meters, great seasonal differences in light intensity and an extensive water transport through the area (Skjoldal et al., 2004b) provides suitable conditions for plankton production during spring and summer (Rey, 2004). Several large pelagic fish stocks use the Norwegian Sea as a feeding area during the spring and summer (Zilanov, 1968; Dragesund et al., 1980; Belikov et al., 1998; Monstad et al., 1998; Holst et al., 2002; Iversen, 2004), where Norwegian Spring Spawning Herring (NSS-herring) (Clupea harengus), Northeast Atlantic mackerel (Scomber scombrus) and blue whiting (Micromesistius poutassou) are the most important species. They feed intensively on the abundant zooplankton in the area, with Calanus finmarchicus as the most important prey (Dalpadado et al., 2000; Gislason and Astthorsson, 2002; Prokopchuk and Sentyabov, 2006). C. finmarchicus is also preyed upon by other types of zooplankton such as krill and amphipods (Melle et al., 2004), which again are preyed upon by pelagic fish themselves. Since 1995 the zooplankton abundance has had a declining trend (Anon, 2009), while the biomass of planktivorous fish in the Norwegian Sea has increased


62

Figure 1. The Norwegian Sea area and surrounding waters. The arrows show the direction of the main currents in the area, where red arrows are warm Atlantic water and blue arrows are cold Arctic water (edited from Loeng and Drinkwater, 2007). Proper management of the Norwegian Sea ecosystem is important for several reasons. First of all, a range of top predators are dependent on the fish and zooplankton species located in the Norwegian Sea, including many species of seabirds (Anker-Nilsen and Lorentsen, 2004) and sea mammals (Nøttestad and Olsen, 2004). Secondly, the area supports a huge fishery where NSS-herring and mackerel are important species, due to their high abundance and financial importance. Thirdly, water flowing out of the Norwegian Sea exports nutrients and plankton into surrounding areas such as the North Sea and the Barents Sea, which is very important for these ecosystems (Aksnes and Blindheim, 1996; Heath et al., 1999; Sundby, 2000; Edvardsen et al., 2003). There has been a great focus on the Norwegian Sea through continuous monitoring of the ecosystem with annual internationally organized surveys, and by extensive research during the Norwegian Mare Cognitum program in the 1990s (Skjoldal et al., 1993b; Skjoldal et al., 2004b). However, major issues are still unclear, such as the cause for the substantial decline in zooplankton observed the last years (Anon, 2009), and why planktivorous fish change their feeding migrations between years (Holst et al., 2002; Utne et al., submitted). To understand how natural environmental oscillations or climate changes affects ecosystems and fish productions, it is necessary to develop suitable approaches to manage marine resources (Travers et al., 2007; Cury et al., 2008). It is not yet clear if the Norwegian Sea ecosystem is regulated by bottom-up and/or top-


63

down processes. Given the complexity of the ecosystem, it is difficult to address the important mechanisms regulating the ecosystem by field studies alone. Models with detailed spatial and temporal resolution are well suited for this purpose. Attempts to calculate the predation pressure from feeding planktivorous fish on zooplankton, and especially on C. finmarchicus, has been done with similar results (Dommasnes et al., 2004; Skjoldal et al., 2004a; Varpe et al., 2005). However, these estimates are ignoring the spatial resolution and need to make several assumptions about important parameters, such as the energetic cost for metabolisms. An alternative method to calculate the predatory impact on zooplankton is by using individual based models (IBMs). These models take into account the spatial heterogeneity and temporal scale (DeAngelis and Gross, 1992; Huse et al., 2002) when calculating the interactions between planktivorous fish and zooplankton. IBMs have been used to model the behavioral response of fish as a function of prey density (e.g. Railsback et al., 1999; Okunishi et al., 2009). IBMs are considered to be a well suited for end-to-end modeling, as they easily combine lower trophic levels (LTL) with higher trophic levels (HTL) at a temporal and spatial scale (Rose et al., 2010). Coupled models Much of the focus so far on ecosystem models using the individual based approach has been on coupled nutrient-phytoplankton-zooplankton (NPZ) models (Rose et al., 2010), also called models of lower trophic levels (LTL). So far, there has been put little effort into coupling models of LTL to higher trophic levels (HTL) by the scientific community, probably due to difficulties with linking spatial and temporal scales of the different models. An example of a successful coupled model is the NEMURO LTL model (Kishi et al., 2007) coupled with the HTL model NEMURO.FISH (Megrey et al., 2007). However, the NEMURO.FISH model system does not take into account spatial detail or variation among individuals. Exploitative competition in advective marine systems is complex since it involves both interactions when species overlap in space and time as well as exploitation of the same resource, but in different space and/or time. In this paper, we present a coupled biophysical model of the Norwegian Sea ecosystem integrating an ocean model, a phytoplankton model and IBMs for Calanus finmarchicus and 3 planktivorous fish stocks. The model is two-way coupled where the different trophic levels provide prey densities for the trophic level above, and feeding at the higher trophic level is used to calculate mortality on the trophic levels below. This model system without planktivorous fish on the top of the food chain is described in this issue (Hjøllo et al., submitted). Here we add IBMs of NSS herring, NEA mackerel and blue whiting migrations to develop the model system further. Objectives Our first objective is to quantify the amount of zooplankton consumed by NSS herring, NEA mackerel and blue whiting in the Norwegian Sea during one year, with a special focus on C. finmarchicus. To achieve this, we adopt fish migration parameters (direction and speed) from a previous study (Utne and Huse, submitted) that used IBMs to recreate historic fish migrations from 1995 to 2003. By using bioenergetics models of fish growth, the total amount of zooplankton eaten by the fish can be estimated. The stock sizes and migration patterns of the fish in 1997 are used to create horizontally resolved fish predation pressure on the zooplankton. The effect on the C. finmarchicus abundance is evaluated, both on a spatial and annual scale. Furthermore, the interspecific competition between the pelagic fish stocks is evaluated by only including one species at the time in the model, and then investigate how the consumption changes in absence of interspecific competition.


64

The modelling approach The purpose of the model is to understand the spatial and trophic interactions between NSS herring, NEA mackerel and blue whiting stocks in the Norwegian Sea. The model system consists of a Eularian biogeochemical ocean model NORWECOM. The ocean model is run offline, while the primary production, zooplankton and fish models are run fully two-way coupled, i.e. with prey densities and mortality rates exchanged on a hourly (primary production and zooplankton) or daily basis. More details on the model can be found in Utne et al (2012). Results Daily consumption and percentage of Calanus in the diet The percentage of C. finmarchicus in the herring, blue whiting and mackerel diet has earlier been calculated from stomach samples taken during surveys in the Norwegian Sea (Dalpadado et al., 2000; Gislason and Astthorsson, 2002; Prokopchuk and Sentyabov, 2006). Here we used unpublished stomach samples to calculate the average percentage of C. finmarchicus in the diet (Fig. 2). These results agree well with the earlier findings. To calculate the total consumption of different prey species throughout the entire feeding season is complicated since the predator diet changes with prey availability. In earlier estimates of consumption of zooplankton by pelagic fish in the Norwegian Sea, has it been assumed that stomach samples taken during May – July are representative for the whole feeding season (Dommasnes et al., 2004; Skjoldal et al., 2004a; Varpe et al., 2005). Our results show that the planktivorous fish have huge changes in their diet composition during their feeding seasons. This is most apparent for mackerel and blue whiting, as C. finmarchicus is only available during parts of their feeding season. Blue whiting generally do not feed on C. finmarchicus from October to April, while mackerel feeds mainly on other prey than C. finmarchicus in October and November. Estimating the percentage of C. finmarchicus in the blue whiting diet is challenging since the juveniles focus their feeding more on C. finmarchicus than the adults do (Bjelland and Monstad, 1997). The blue whiting stomach samples we use to validate the percentage of C. finmarchicus in the diet were not separated into adults and juveniles. Skjoldal et al (2004a) estimated blue whiting to have around 10% of C. finmarchicus in their diet during the summer, but mentioned that this was probably an underestimation since mainly adults were sampled in the analyses. This is slightly lower than our results, where the adult blue whiting diet consists of 15-25% C. finmarchicus during the summer. Adult blue whiting have in average stored energy reserves corresponding to 19% of their body weight at the end of the year. This is higher than for mackerel and herring, and is maybe too high. Most of the adults are two years old, and since they have a somatic growth of more than 5% year-1 (ICES, 2008a) more of the energy reserves should have been allocated to body weight for blue whiting. Nevertheless, this does not affect the overall results. The consumption estimated by the model (Table 1) is 35 mill. Tonnnes of C. finmarchicus and 47 mill. Tonnes of other prey. These estimates are based model simulations where the three stocks compete for food. When each of the stocks are simulated alone without the competition from the two others, the consumption of C. finmarchicus is increased for all three stocks (Table 2). While the incrase is modes for the herring and mackerel, the increase is more than 50% for the blue whiting, indicating that the blue whiting is suffering from the exploitative competition from herring and mackerel.


65

Perc

enta

ge C

fin

marc

hic

us in t

he d

iet

Jan Mar May Jul Sep Nov Jan

010

30

50

70

90 a)

Perc

enta

ge C

fin

marc

hic

us in t

he d

iet


010

30

50

70

90 b)

Perc

enta

ge C

fin

marc

hic

us in t

he d

iet


010

30

50

70

90 c)

Perc

enta

ge C

fin

marc

hic

us in t

he d

iet


010

30

50

70

90 d)

Figure 2. The percentage of C. finmarchicus in the diet throughout the year, a) herring b) mackerel c) juvenile blue whiting d) adult blue whiting. The horizontal line shows the average percentage of C. finmarchicus in the diet from the stomach samples. Predation effect on zooplankton abundance and its spatial distribution The fish model can be seen as a dynamic extension of the uniform fish mortality applied in Hjøllo et al(Hjøllo et al., 2012), and we here compare these two C. finmarchicus mortality approaches. With the uniform fish mortality, fish predation pressure is parameterized as daylight and prey size dependent mortality rate, while with the dynamic fish mortality, mortality is regionally varying as specified by the migration patterns described in (Utne and Huse, 2012).

The spatial distribution of C. finmarchicus in the upper 400m of the water column in Mid-June show that with the uniform mortality, high densities of C. finmarchicus can be found along the Norwegian coast and in the Barents Sea entrance (Fig. 3). With the dynamic fish model, the Barents Sea opening abundance is increased , while along the Norwegian coast the abundance is lower, in accordance with the lack of (Barents Sea opening) or presence of (along the coast) fish predation in


66

the dynamic fish model. For mid-August, the same situation is seen in south: the dynamic fish distribution results in higher abundance near Iceland, reflecting lower fish predation there. In the deeper basins of the Norwegian Sea, the fish distribution from the dynamic fish mortality model seems to give somewhat elevated abundance in Mid-June, while depletion is seen in August. Table 1. Estimated total consumption of Calanus finmarchicus and other zooplankton in 1997 (in million tons).

C. finmarchicus Other prey

Herring 24.5 26 Blue whiting 4 14 Mackerel 6.5 7

Total 35 47

Table 2. Difference in consumption when only one of the species are included in the model at once, compared to when all species are included in the model simultaneously.

Total consumption Consumption C. finmarchicus

Herring 101.99 % 110.15 % Blue whiting 104.71 % 152.32 %

Mackerel 104.32 % 112.35 %

Figure 3. The spatial distribution of C. finmarchicus in mid-June and mid-August from the model system when the predation from fish is uniform over the entire domain (a and c) and when the


67

predation from fish is spatially explicit from the fish model (b and d). DISCUSSION We estimated that herring, mackerel and blue whiting consumed altogether 82 million tons zooplankton, including 35 million tons C. finmarchicus. This gives a consumption/biomass ratio in the range 5.2-6.3. Herring is the most important species, with an annual consumption of more than 50 million tons zooplankton and nearly 50% of this is C. finmarchicus. Blue whiting is the species mostly affected by the interspecific competition, as blue whiting had the largest increase in consumption when the two other species were not included in the model. C. finmarchicus is experiencing a high predation pressure from the pelagic fish species in the central Norwegian Sea. There are high densities of C. finmarchicus in the surrounding areas where pelagic fish is absent. Predation pressure by planktivorous fish The estimates of total consumption of zooplankton by planktivorous fish in the Norwegian Sea agrees reasonable well with previous estimates (Dommasnes et al., 2004; Skjoldal et al., 2004a; Varpe et al., 2005). Dommasnes et al (2004) estimated herring to have a consumption/biomass ratio of 4.5. Skjoldal et al (2004a) estimated 8 million tons herring, 4 million tons blue whiting and 4 million tons mackerel to consume 60 million tons zooplankton year-1. Of this was 43.5 million tons defined as primary carnivore species, which is mainly C. finmarchicus. Varpe et al (2005) estimated the average herring stock between 1990-2003 to consume a total of 38 million tons zooplankton per year. The first two papers are based on consumption/biomass estimates or estimates of production/biomass as a function of organism size. The last paper uses the same bioenergetics model for fish growth as we use in this paper. The main difference is the swimming speed and water temperature, which are equal for all individuals in Varpe et al (2005). A weakness with that study was that increase in herring energy density (J g fish-1) during the feeding season is not properly handled in the model. The estimated result of 34-37% individual growth during the feeding season are ignoring the prey needed for the increased energy content of one gram herring during this period, which is increasing by around 100 % (Slotte, 1999). With a swimming speed of one bl s-1 in the model, the estimate of 38 mill tonnes of zooplankton consumed by 6.1 million tonnes of herring does not agree with the results from the present study. With an annual consumption of 82 million tons of zooplankton, it is clear that the planktivorous fish has a larger impact on the zooplankton community in the Norwegian Sea than earlier assumed. The annual consumption/biomass ratio of 5.2-6.3 is within the range of 3-7 estimated for herring in several independent studies summarized in Dommasnes et al (2004). In the model the consumption of other prey exceeded the consumption of C. finmarchicus. Predation by planktivorous fish has earlier been considered low compared to amphipods (Themisto) and Gonatus (Skjoldal et al., 2004a). This has lead to the belief that planktivorous fish has a low impact on the zooplankton abundance compared to Gonatus, krill and amphipods. As the abundance of pelagic fish increased after 1997, and the biomass of blue whiting and herring was more than 50% higher in 2006, it is clear that the pelagic fish has a large impact on the zooplankton abundance in the Norwegian Sea. Daily consumption and percentage of Calanus in the diet Stomach fullness has been very low in several diet studies, indicating a low feeding activity even during the main feeding season (Dalpadado et al., 2000; Gislason and Astthorsson, 2002; Prokopchuk and Sentyabov, 2006). The calculated prey consumption in the model results in a high


68

daily stomach fullness, with a daily consumption around 4% of the fish weight. In the model we assume that total digestion time is maximum 24 hours. Young herring feeding in the Barents Sea has been reported to have an average stomach weight as high as 5.4 % of the fish weight (Godiksen et al., 2006), which is higher than our daily estimates for adult fish. Fish larva is digested very rapidly, with a digestive rate of 0.3 h-1 for capelin larvae in herring stomachs (Hallfredsson et al., 2007). The digestive rate of zooplankton is poorly studied, but it will take longer time to digest zooplankton due to its hard external shell. Darbyson et al (2003) estimated herring feeding mainly on copepods and krill to digest the prey in less than 8 hours. The daily consumption estimates in the presented model is thereby well within reasonable limits as the fish has no problem digesting the prey during 24 hours. The percentage of C. finmarchicus in the herring, blue whiting and mackerel diet has earlier been calculated from stomach samples taken during surveys in the Norwegian Sea (Dalpadado et al., 2000; Gislason and Astthorsson, 2002; Prokopchuk and Sentyabov, 2006). Here we used unpublished stomach samples to calculate the average percentage of C. finmarchicus in the diet. These results agree well with the earlier findings. To calculate the total consumption of different prey species throughout the entire feeding season is complicated since the predator diet changes with prey availability. In earlier estimates of consumption of zooplankton by pelagic fish in the Norwegian Sea, has it been assumed that stomach samples taken during May – July are representative for the whole feeding season (Dommasnes et al., 2004; Skjoldal et al., 2004a; Varpe et al., 2005). Our results show that the planktivorous fish have huge changes in their diet composition during their feeding seasons. This is most apparent for mackerel and blue whiting, as C. finmarchicus is only available during parts of their feeding season. Blue whiting generally do not feed on C. finmarchicus from October to April, while mackerel feeds mainly on other prey than C. finmarchicus in October and November. Estimating the percentage of C. finmarchicus in the blue whiting diet is challenging since the juveniles focus their feeding more on C. finmarchicus than the adults do (Bjelland and Monstad, 1997). The blue whiting stomach samples we use to validate the percentage of C. finmarchicus in the diet were not separated into adults and juveniles. Skjoldal et al (2004a) estimated blue whiting to have around 10% of C. finmarchicus in their diet during the summer, but mentioned that this was probably an underestimation since mainly adults were sampled in the analyses. This is slightly lower than our results, where the adult blue whiting diet consists of 15-25% C. finmarchicus during the summer. Adult blue whiting have in average stored energy reserves corresponding to 19% of their body weight at the end of the year. This is higher than for mackerel and herring, and is maybe too high. Most of the adults are two years old, and since they have a somatic growth of more than 5% year-1 (ICES, 2008a) more of the energy reserves should have been allocated to body weight for blue whiting. Nevertheless, this does not affect the overall results. Predation effect on C. finmarchicus abundance and its spatial distribution The predation pressure from the fish is spatially and time span restricted to the areas where fish is present, but the effect on C. finmarchicus abundance is dependent also on advective processes and interaction with the available grazing fields. Surveys in May where WP2 nets were used showed that there were lower zooplankton densities in 1997 than in 1995 and 1996. The pattern was however the same in all three years with the highest densities in the northwestern part of the Norwegian Sea and the lowest densities in the southern parts (ICES, 1997). The same pattern is emerging from the C finmarchicus model which shows that the fish predation from the dynamic model results in a horizontal distribution of C finmarchicus similar to the survey data.


69

Compared to the uniform fish predation, the total fish predation is reduced by ~50%. In the central Norwegian Sea, where the fish predation is highest, the C. finmarchicus biomass is still weakly increased with the fish model. The predation pressure is lower in the surrounding areas further north in June, and further south and east in August, and more C. finmarchicus suffer from food limitations. The C. finmarchicus abundance in areas without direct predation from fish is affected by surrounding areas, as advection from nearby areas, but their relative importance is not readily seen. The present simulation addressed the seasonal development in the Norwegian Sea over one year only. In future studies, it will be interesting to perform multiple year simulations where the effect of interannual variable fish mortality through variations in fish migration pattern can be studied. Interspecific competition Blue whiting had the largest increase, both in total consumption and in consumption of C. finmarchicus when the other species were absent. This shows that blue whiting is the species most affected by interspecific competition. While herring and mackerel are fast swimming species mostly located in schools, blue whiting is smaller, has a lower swimming speed and are located in clusters instead of schools (Skjoldal et al., 2004b). The herring enter the feeding area before the main components of the other species and this is probably the cause for the small increase in consumption when the other species were absent. Blue whiting are probably more dependent on the density of prey surrounding the fish, than herring and mackerel that actively search for areas with high densities of prey. The blue whiting stock has declined in the last five years, mainly due to poor recruitment (ICES, 2009). The mackerel and herring stocks are still large and in good shape (ICES, 2008b, 2009). This also indicates that blue whiting is the poorest competitor of the three species, especially when competing with other species for C. finmarchicus. Model uncertainty The bioenergetics model is sensitive to respiration (Varpe et al., 2005; Megrey et al., 2007). The respiration cost in bioenergetics model is based on experiments where single fish are swimming against currents in a tank (Jobling, 1994). This is an unnatural condition for wild migrating fish, which benefit from the aerodynamic advantages of schooling (Pitcher, 1986). The endurance of fish may increase two to six times when schooling (Weihs, 1973). Wild fish may also benefit from vertical movements making it possible to take advantage of currents at different depths moving in a favorable direction for the fish, which is shown with tidal currents (Gibson, 2003). The energetic cost for respiration is therefore uncertain, and it is possible that the fish swimming speed should have been changed. The swimming speed in the model is lower than in situ estimates during surveys (Misund et al., 1998; Kvamme et al., 2003). This adds uncertainty to the model results. Since the fish model is a numerical model, there is a challenge with superindividuals within the same square experiencing different prey densities at the same time step. This is caused by fish superindividuals feeding one at the time within the same time step, causing less available prey for the ones feeding last. This was overcome with the two step procedure introduced by Huse et al. (2004), where the individuals within the same square at a given time step will have the same feeding intensity (g-1 g fish). This approach reduces the individual variability since all individuals in the same area have the same feeding success. An alternative approach is to let the superindividuals feed in a random order. With this approach there is no unrealistic constraint on the feeding process but at the cost of a larger variation in energy intake between individuals within the same area and time step. Random selection can also be used to decide the species feeding order during each time


70

step. The fish were not allowed to consume more than 80 % of C finmarchicus biomass in each square during each time step. The possibility of the fish grazing down the zooplankton during one time step can be avoided by reducing the length of the time step in the model. This would however increase the simulation time which is already very high even with a time step of 24 hours. In this fish model a constant density of other prey than C. finmarchicus was present independent of time and spatial position. Whether the fish focused their feeding on other prey organisms depends on the density of C. finmarchicus. This is a simple approach to include other prey in the model for which there exists little knowledge of the abundance and spatial dynamics. An alternative approach is to assume that the fish eat a fixed amount of other prey throughout the year, and the seasonal feeding on C. finmarchicus comes in addition to this fixed amount other prey (Huse et al., 2004). In the presented model the feeding intensity is relatively high and stable throughout the feeding season for all species. In addition the water temperature affects the maximum energy intake and the energy cost associated with respiration. This leads for instance to a higher feeding intensity for herring during April than in May and June, a result that contradicts earlier findings (Dalpadado et al., 2000). End-to-end modeling of ecosystem dynamics We have used a fully coupled model system with LTL and HTL to model the annual predation pressure of herring, blue whiting and mackerel on C. finmarchicus in the Norwegian Sea. Whether large marine ecosystems are top-down or bottom-up regulated have received great attention the latest years (Cury et al., 2000; Ware and Thomson, 2005). This question has also been raised for the Norwegian Sea ecosystem, but no answer has been given yet. The presented model system is ideal for studying ecosystem regulations and how the key species in the ecosystem respond to different environmental conditions. The coupled model system bears similarities to the coupled model system combining the LTL NEMURO (Kishi et al., 2007) with the HTL NEMURO.FISH (Megrey et al., 2007). A major difference in the presented model compared to NEMURO and NEMURO.FISH is the spatial resolution and individual variability. In the presented model the fish experience spatially explicit changes in prey availability, currents and water temperature, which is one of the advantages with IBMs. IBMs should be complex enough to recreate the modeled pattern with fewest possible input parameters (Grimm and Railsback, 2005). When reducing model complexity by using homogenously distributed input parameters, important emerging properties may be lost in the simulations. Further can the full web structure and changing importance of mechanisms be lost (Fulton et al., 2004). The presented model can potentially be used to forecast how the ecosystem will react to various climate scenarios. It is also possible to study potential changes in the ecosystem due to the predation pressure on C finmarchicus executed by the planktivorous fish. By running the model for several consecutive years with changing stock populations initiated from VPA data, it is possible to test if there are top-down regulations affecting the abundance of C finmarchicus, and if this affects the ecosystem in general. The effect of changing temperature and hydrographic conditions and bottom-up effects may also be estimated. We believe that the individual based approach is well adapted for future ecosystem studies, in spite of the low complexity of these models. Conclusions We have presented a fully coupled model system with planktivorous fish representing the HTL. The migration patterns and the spatial distribution of the fish are driven by historic survey observations and by the zooplankton densities the fish encounter. Bioenergetics models were used to estimate


71

the total consumption of zooplankton by herring, blue whiting and mackerel in 1997, which was 82 mill tons. The annual consumption of C. finmarchicus was calculated to be 35 mill tons. This predation pressure from the pelagic fish stocks did not reduce the abundance of C. finmarchicus in the next generation, but had a large effect on the spatial distribution of C. finmarchicus. In the area with the most intense feeding pressure, mainly in the Atlantic water masses to the west of central Norway, the C. finmarchicus abundance was depleted. South and north of the areas with the main fish predation were there high abundance of C. finmarchicus after the main feeding period for fish. Herring was the species that was least affected by the interspecific competition with the other species, while blue whiting was the species suffering most from interspecific competition.


72

8 Effects of interactions between fish populations on ecosystem dynamics in the Norwegian Sea Geir Huse, Jens Christian Holst, Kjell Utne, Leif Nøttestad, Webjørn Melle, Aril Slotte, Geir Ottersen The Norwegian Sea (NS) is the feeding ground for some of the largest fish stocks in the world, including Norwegian spring spawning (NSS) herring (Clupea harengus Linnaeus, 1758), blue whiting (Micromesistius poutassou Risso, 1827) and the Northeast Atlantic (NA) mackerel (Scomber scombrus Linnaeus, 1758). These planktivorous stocks have substantial spatial and dietary overlap (e.g. Nøttestad et al., 1997; Dalpadado et al., 2000; Kaartvedt, 2000), and are often collectively referred to as the “pelagic complex” in the Norwegian Sea. Due to their high abundances, they can potentially have a strong ecological impact on the ecosystem and each other (Skjoldal et al., 2004c). The NSS herring collapsed in the late 1960’s and rebuilt during the 1980’s (Dragesund et al., 1997). Following the herring collapse, high abundances of blue whiting were discovered in the Norwegian Sea (Misund et al., 1998), and it has been speculated that the blue whiting population increased concurrently with the collapse of the NSS herring (Skjoldal et al., 1993a), but the evidence remains inconclusive (Daan, 1980). Since the late 1980’s the abundance of fish in the NS has increased steadily and this has increased the potential for interactions between the planktivorous stocks (Figure 1). This has lead to the proposition of the hypothesis that the planktivorous fish populations feeding in the NS have interactions that negatively affect individual growth, mediated through depletion of their common zooplankton resource. In the recent decade the trend of a decreasing zooplankton biomass in the NS continued and the biomass now remains low (Figure 1). The fish biomass peaked in 2004 and has since decreased somewhat, but remains fairly high. The abundance of blue whiting increased until 2004, and the range of the horizontal distribution expanded in a north-westerly direction during this period. Strong year classes of mackerel from 2001 and 2002, together with increasing temperatures, resulted in an increased number of mackerel in the Norwegian Sea (Payne et al. 2012; Utne et al. 2012a). Furthermore, there were rather substantial changes in the migration pattern of herring during the study period and thus high inter-annual variability in horizontal overlap between the species. There was a relatively high spatial overlap between the species during the 1990s, with a southern centre of gravity (for all three species), but due to the northern displacement of especially herring and blue whiting, the overlap decreased in the early 2000s. As mackerel stayed mainly

south of 70N and NSS herring north of 70N, the horizontal overlap between these species was limited (Utne et al. 2012a; Utne & Huse 2012). The horizontal overlap between blue whiting and mackerel was extensive in some years, but since the blue whiting prefers deeper water than mackerel, the vertical overlap is low. There was pronounced inter-annual variability in the vertical distribution of blue whiting and herring (Huse et al. 2012). The vertical distribution appears to be linked so that herring occurs shallower when the abundance or overlap with blue whiting is high. This indicates that there is interaction between those species. The diet of the three species varies between years and with season. The peak feeding season for herring and blue whiting is typically in May-June, whereas for mackerel it is in July. The herring diet is dominated by Calanus finmarchicus (Gunnerus, 1770), particularly early in the feeding season. Later in the season the diet is more varied and less dependent on C. finmarchicus (Langøy et al.


73

2012; Utne et al. 2012b). Krill and amphipods then become more important as prey. The blue whiting has an ontogenetic shift in its diet associated with the move into deeper waters at an age of around two years. The diet shift is characterized by going from a C. finmarchicus dominated diet to a diet dominated by krill and amphipods. Mackerel seems to be more opportunistic and adjust to prey availability more than the other two fish species, but the diet is often dominated by C. finmarchicus (Langøy et al. 2012). Due to the dynamic space uses by pelagic fish, one needs to capture their 3D spatial distribution in order to study their interactions. The role of space in ecology remains elusive, and the subject has been referred to as the “final frontier for ecological theory” (Kareiva, 1994). Individual based models (IBM’s) with super-individuals have been developed for the copepod C. finmarchicus, the main meso-zooplankton component of the NS, and for the NSS herring, blue whiting, and NA mackerel (Hjøllo et al. 2012, Utne et al. 2012b). These models are coupled with the biogeochemical model NORWECOM and the Regional Ocean Model System (ROMS). The result is a unique ecosystem model complex that integrates ocean physics, feeding, growth, fine-scale movement and life history traits of key plankton and fish species with full feedback of energy between different trophic levels. The model system has been developed and is used to simulate the spatial overlap between the stocks on a daily fine scale basis. Whereas the data analysis discussed above provides snapshots of the distributions, the fish IBM has a daily time step and provides daily predictions of overlap. The migration model illustrates that the overlap between the species is highly dynamic within the season and varies between years (Utne & Huse 2012). Preliminary model simulations using the fully coupled model system suggest that the planktivorous stocks exert a considerable predation pressure on the zooplankton resources in the Norwegian Sea (Utne et al. 2012b). During the 1980’s the biomass of the planktivorous fish stocks was about a third of the peak biomass in 2004, and the biomass of zooplankton was much higher (Figure 1). The ratio of estimated fish consumption to production ratio for C. finmarchicus has therefore been high in recent years and indicates that fish predation has had an increasing impact on the C. finmarchicus population (Figure 1). There are uncertainties in the absolute levels of this ratio, but the substantial increase in the recent decade makes it plausible that the reduction in zooplankton biomass seen after 2002 is caused by fish predation. In 1997 there was a very low plankton index value (Figure 1) that was probably not attributed to fish predation, but rather to unfavourable conditions for primary and/or secondary production. The migration pattern of the fish has changed to become extended further to the west during the period and the former main feeding areas in the central Norwegian Sea have virtually been abandoned. The NSS herring has had a downward trend in length at age over time (Figure 2). In recent years the mackerel length at age has decreased concurrently with an increase in the mackerel stock size (Figure 2). For the blue whiting a decreasing trend in length at age was shifted to a positive trend in 2008 (Figure 2). In order for species interaction to qualify as competition, it has to have a negative impact on at least one of the interacting species. Spearman rank correlations between length at age 6 and stock biomasses are given in Table 1. For herring there are clear signs of both intra- and interspecific competition, while for the mackerel only the intraspecific term is significant (Table 1). For the blue whiting all the three correlations are significant, but here the interspecific relationships are the strongest. This is in line with the simulation results in Utne et al. (2012b) which indicated that the blue whiting foraging rate was negatively affected by the feeding interaction with the other stocks, whereas the other two stocks were much less affected. The recent increase in the blue whiting length at age is probably due to very low intraspecific competition at the present low stock biomass. So even though the size


74

at age is likely to depend on climatic conditions (Holst et al., 2004), light regime (Varpe & Fiksen 2010), and nursery area (Holst and Slotte, 1998) among other factors, there are clear signs of intra- and interspecific competition in the pelagic complex. The planktivorous fishes also feed on krill, amphipods and mesopelagic fish, which also are predators on Calanus. It is a bit paradoxical that in spite of the observed reduction in zooplankton biomass, an increased abundance of planktivorous fish may have decreased the Calanus predation by macroplankton and mesopelagic fish and thus increased the carrying capacity for planktivorous fish (Skjoldal et al 2004b). In order to address how local depletion of zooplankton due to fish predation might be dispersed over the winter (Olsen et al., 2007), the drift patterns of RAHFOS floats drifting at 800 m in the Norwegian Sea over the winter were studied (Søiland & Huse 2012). This depth is representative of the depth where the C. finmarchicus drift passively in diapause during winter. The drift trajectories of the RAHFOS floats varied substantially in displacement direction and magnitude. The results show that the transport of overwintering C. finmarchicus is likely to disperse the distribution and substantially diminish “traces” of feeding by the planktivorous fish.

In conclusion there are rather strong overall spatial interactions within the pelagic complex during the feeding season in the NS, with considerable potential for exploitative competition for common zooplankton resources due to the large overlap in diet. During the study period there has been a strong build-up of biomass of planktivorous fish in the Norwegian Sea. The negative relationships between length at age and stock biomass, the pronounced reduction in zooplankton abundance witnessed in the Norwegian Sea in recent years, and expansion in spatial distribution of fish indicate that the biomass of planktivorous fish in the area has been above the carrying capacity. All the stocks showed signs of density dependent length growth, whereas for herring and blue whiting there were also significant effects of interspecific competition (Table 1). Since 2009 there has been an upward trend in the zooplankton biomass (Figure 3). At the same time there has been a pronounced decrease in the herring biomass due to poor recruitment to the stock since 2004. Nevertheless the total biomass of pelagic fish in the Norwegian Sea has remained stable due to increases in the mackerel and blue whiting stocks (Figure 3). Herring is generally the biggest zooplankton consumer in the Norwegian Sea. Furthermore the herring enters the Norwegian Sea in March-April, much earlier than the other stocks. It can therefore be argued that the herring has a greater predatory effect on the zooplankton since the herring feed on the the ascending generation of Calanus finmarchicus before the main reproduction of Calanus has started. This early predation therefore is likely to have a greater effect on the Calanus dynamics than predation in late summer when much of the Calanus has descended to overwintering at depth. In this manner the Calanus is experiencing predation release with the recent reduction in the herring stock. The increase in zooplankton may also be related to favourable physical and phytoplankton conditions, although the available data do not imply a strong bottom up impact. The results therefore support the hypothesis that the planktivorous fish populations feeding in the NS have interactions that negatively affect individual growth, mediated through depletion of their common zooplankton resource. It will be important to include these findings in the future ecosystem based management of the Norwegian Sea.


75

Table 1. Spearman rank correlations (p-values) between fish length at age 6 against intraspecific (single) stock biomass, interspecific biomass (sum of two other stocks), and total fish biomass (sum of all three). For herring the spawning stock biomass was used and for mackerel and blue whiting total stock biomass (1+) was used. Data from the period 1982-2011 were used and the length data were taken from the IMR data base and the biomass data for the period 1982-1987 were taken from ICES (2007) and for the period 1988-2011 were taken from ICES (2011).

Herring Mackerel Blue whiting

Intraspecific biomass 5.36E-08 0.04889 0.01359

Interspecific biomass 0.004782 0.5869 0.001830

Total biomass 4.96E-05 0.4618 0.0001404

0

0.2

0.4

0.6

0.8

1

0

2

4

6

8

10

12

14

16

18

19

88

19

90

19

92

19

94

19

96

19

98

20

00

20

02

20

04

20

06

20

08

20

10

20

12

Co

nsu

mp

tio

n/p

rod

uct

ion

rati

o

Bio

ma

ss o

f p

ela

gic

fish

(mill

. to

nn

es)

Z

oo

pla

nk

ton

ind

ex

Year

Mackerel Blue Whiting Herring

Zoo. index Estimated C/P ratio Modelled C/P ratio

From Huse & al. 2012

Plankton production and fish consumption

”Bottom up” ”Top down”

Figure 1. Developments in fish biomass, plankton biomass and ratio of estimated fish consumption of Calanus finmarchicus divided by the estimated C. finmarchicus production for copepodite stage 4 to adult. The annual consumption estimates are based on consumption/biomass relations and diet data in Skjoldal et al. (2004a), and the annual production estimates are based on the total Calanus production estimate in Skjoldal (2004a) divided by stage using data from Aksnes & Blindheim (1996) multiplied by the Zooplankton index relative to the maximum value.


76

Figure 2. Average length at age 6 (±95% confidence intervals) for NSS herring, blue whiting and NA mackerel.


77

0

2

4

6

8

10

12

14

16

0

2

4

6

8

10

12

14

16

1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014

Pla

nkt

on

bio

mas

s (g

m-2

)

Fish

bio

mas

s (m

ill. t

on

ne

s)

Year

Herring Mac Blue whiting Total fish abundance Plankton

Figure 3. Zooplankton and fish biomass in the Norwegian Sea during 1988 to 2014. References Aksnes, D.L., Blindheim, J., 1996. Circulation patterns in the North Atlantic and possible impact on

population dynamics of Calanus finmarchicus. Ophelia 44, 7-28. Anker-Nilsen, T., Lorentsen, S.H., 2004. Seabirds in the Norwegian Sea., in: Skjoldal, H.R. (ed.), The

Norwegian Sea Ecosystem. Tapir Academic Press, Trondheim, Norway, pp. 435-444. Anon, 2009. Havets Ressurs og miljø. Fisken og Havet 1, 1-185. Belikov, S., Jakupsstovu, S.H., Shamrai, E., Thomsen, B., 1998. Migration of mackerel during summer

in the Norwegian Sea. ICES C.M.1998/AA:8 1-14. Bjelland, O., Monstad, T., 1997. Blue whiting in the Norwegian Sea, spring and summer 1995 and

1996. ICES C.M.1997/CC 15, 1-25. Cury, P., Bakun, A., Crawford, R.J.M., Jarre, A., Quinones, R.A., Shannon, L.J., Verheye, H.M., 2000.

Small pelagics in upwelling systems: patterns of interaction and structural changes in "wasp-waist" ecosystems. ICES J. Mar. Sci. 57, 603-618.

Cury, P.M., Shin, Y.J., Planque, B., Durant, J.M., Fromentin, J.M., Kramer-Schadt, S., Stenseth, N.C., Travers, M., Grimm, V., 2008. Ecosystem oceanography for global change in fisheries. Trends Ecol. Evol. 23, 338-346.

Daan, N., 1980. A review of replacement of depleted stocks by other species and the mechanisms underlying such replacement. Rapp Proc Verb Réun Cons Int l'Exp Mer 177, 405-421.

Dalpadado, P., Ellertsen, B., Melle, W., Dommasnes, A., 2000. Food and feeding conditions of Norwegian spring-spawning herring (Clupea harengus) through its feeding migrations. ICES J. Mar. Sci. 57, 843-857.

Darbyson, E., Swain, D.P., Chabot, D., Castonguay, M., 2003. Diel variation in feeding rate and prey composition of herring and mackerel in the southern Gulf of St Lawrence. Journal of Fish Biology 63, 1235-1257.

DeAngelis, D.L., Gross, L.J., 1992. Individual-based Models and Approaches in Ecology. Chapman & Hall, London 525 pp.


78

Dommasnes, A., Melle, W., Dalpadado, P., Ellertsen, B., 2004. Herring as a major consumer in the Norwegian Sea. ICES J. Mar. Sci. 61, 739-751.

Dragesund, O., Hamre, J., Ulltang, Ø., 1980. Biology and population dynamics of Norwegian spring-spawning herring. Rapp Proc Verb Réun Cons Int l'Exp Mer 177, 43-71.

Dragesund, O., Johannessen, A., Ulltang, Ø., 1997. Variation in migration and abundance of Norwegian spring spawning herring (Clupea harengus L.). Sarsia 82, 97-105.

Edvardsen, A., Slagstad, D., Tande, K.S., Jaccard, P., 2003. Assessing zooplankton advection in the Barents Sea using underway measurements and modelling. Fisheries Oceanography 12, 61-74.

Fulton, E.A., Smith, A.D.M., Johnson, C.R., 2004. Effects of spatial resolution on the performance and interpretation of marine ecosystem models. Ecol. Model. 176, 27-42.

Gibson, R.N., 2003. Go with the flow: tidal migration in marine animals. Hydrobiologia 503, 153-161.

Gislason, A., Astthorsson, A.S., 2002. The food of Norwegian spring-spawning herring in the western Norwegian Sea in relation to the annual cycle of zooplankton. Sarsia 87, 236-247.

Godiksen, J.A., Hallfredsson, E.H., Pedersen, T., 2006. Effects of alternative prey on predation intensity from herring Clupea harengus and sandeel Ammodytes marinus on capelin Mallotus villosus larvae in the Barents Sea. J. Fish Biol. 69, 1807-1823.

Grimm, V., Railsback, S.F., 2005. Individual-based modeling and ecology. Princeton University Press, Princeton 428 pp.

Hallfredsson, E.H., Pedersen, T., Praebel, K., 2007. Estimation of digestion rates for herring Clupea harengus L. feeding on fish larvae. J. Fish Biol. 70, 638-643.

Heath, M.R., Backhaus, J.O., Richardson, K., McKenzie, E., Slagstad, D., Beare, D., Dunn, J., Fraser, J.G., Gallego, A., Hainbucher, D., Hay, S., Jonasdottir, S., Madden, H., Mardaljevic, J., Schacht, A., 1999. Climate fluctuations and the spring invasion of the North Sea by Calanus finmarchicus. Fish. Oceanogr. 8, 163-176.

Hjøllo, S.S., Huse, G., Skogen, M.D., Melle, W., 2012. Modelling secondary production in the Norwegian Sea with a fully coupled physical/primary production/individual-based Calanus finmarchicus model system. Mar. Biol. Res. 8, 508-526.

Hjøllo, S.S., Huse, G., Skogen, M.D., Melle, W., submitted. Modeling secondary production in the Norwegian Sea with a fully coupled physical/primary production/Individual-Based Calanus finmarchicus model system. Marine Biology Research.

Holst, J.C., Dragesund, O., Hamre, J., Misund, O.A., Østvedt, O.J., 2002. Fifty years of herring migrations in the Norwegian Sea. ICES Mar. Sci. Symp. 215, 352-360.

Holst, J.C., Røttingen, I., Melle, W., 2004. The Herring, in: Skjoldal, H.R. (ed.), The Norwegian Sea Ecosystem. Tapir Academic Press, Trondheim, Norway, pp. 203-226.

Holst, J.C., Slotte, A., 1998. Effects of juvenile nursery on geographic spawning distribution in Norwegian spring-spawning herring (Clupea harengus L.). ICES J. Mar. Sci. 55, 987-996.

Huse, G., Giske, J., Salvanes, A.G.V., 2002. Individual-based models, in: Hart, P.J.B., Reynolds, J. (eds.), Handbook of Fish and Fisheries, vol. 2. Blackwell Science, Oxford, pp. 228-248.

Huse, G., Johansen, G.O., Gjøsæter, H., Bogstad, B., 2004. Studying spatial and trophic interactions between capelin and cod using individual-based modelling. ICES J. Mar. Sci. 61, 1201-1213.

ICES, 1997. Report on surveys of the distribution abundance and migrations of the Norwegian springspawning herring, other pelagic fish and the environment of the Norwegian Sea and adjacent waters in late winter, spring and summer of 1997. ICES C.M. Y:04, 63pp.

ICES, 2007. Report of the Northern Pelagic and Blue Whiting Fisheries Working Group (WGNPBW)


79

229 pp. ICES, 2008a. Report of the Planning Group on Northeast Atlantic Pelagic Ecosystem Surveys

(PGNAPES), 19–22 August 2008, Hirsthals, Denmark. ICES C.M.2008/RMC:0, 87pp. ICES, 2008b. Report of the Working Group on Widely Distributed Stocks (WGWIDE). ICES CM

2008/ACOM:13, 58-182. ICES, 2009. Report of the Planning Group on Northeast Atlantic Pelagic Ecosystem Surveys

(PGNAPES), 18–21 August 2009, Tórshavn, Faroe Islands. ICES C.M.2009/RMC:6, 139 pp. ICES, 2011. Report of the Working Group on Widely Distributed Stocks (WGWIDE), 23 - 29 August

2011, ICES Headquarters, Copenhagen, Denmark. ICES CM 2011/ACOM:15. 642. Iversen, S.A., 2004. Mackerel and horse mackerel, in: Skjoldal, H.R. (ed.), The Norwegian Sea

ecosystem. Tapir academic press, pp. 289-300. Jobling, M., 1994. Fish Bioenergetics. Chapman and Hall, London., p. 309. Kaartvedt, S., 2000. Life history of Calanus finmarchicus in the Norwegian Sea in relation to

planktivorous fish. ICES J. Mar. Sci. 57, 1819-1824. Kareiva, P., 1994. Space: The final frontier for ecological theory. Ecology 75, 1-1. Kishi, M.J., Kashiwai, M., Ware, D.M., Megrey, B.A., Eslinger, D.L., Werner, F.E., Noguchi-Aita, M.,

Azumaya, T., Fujii, M., Hashimoto, S., Huang, D.J., Iizumi, H., Ishida, Y., Kang, S., Kantakov, G.A., Kim, H.C., Komatsu, K., Navrotsky, V.V., Smith, S.L., Tadokoro, K., Tsuda, A., Yamamura, O., Yamanaka, Y., Yokouchi, K., Yoshie, N., Zhang, J., Zuenko, Y.I., Zvalinsky, V.I., 2007. NEMURO - a lower trophic level model for the North Pacific marine ecosystem, pp. 12-25.

Kvamme, C., Nøttestad, L., Ferno, A., Misund, O.A., Dommasnes, A., Axelsen, B.E., Dalpadado, P., Melle, W., 2003. Migration patterns in Norwegian spring-spawning herring: why young fish swim away from the wintering area in late summer. Marine Ecology-Progress Series 247, 197-210.

Loeng, H., Drinkwater, K., 2007. An overview of the ecosystems of the Barents and Norwegian Seas and their response to climate variability. Deep-Sea Research Part Ii-Topical Studies in Oceanography 54, 2478-2500.

Megrey, B.A., Rose, K.A., Klumb, R.A., Hay, D.E., Werner, F.E., Eslinger, D.L., Smith, S.L., 2007. A bioenergetlics-based population dynamics model of Pacific herring (Clupea harengus pallasi) coupled to a lower trophic level nutrient-phytoplankton-zooplankton model: Description, calibration, and sensitivity analysis. Ecol. Model. 202, 144-164.

Melle, V., Ellertsen, B., Skjoldal, H.R., 2004. Zooplankton: The link to higher trophic levels, in: Skjoldal, H.R. (ed.), The Norwegian Sea Ecosystem. Tapir Academic Press, Trondheim, pp. 137-202.

Misund, O.A., Vilhjalmsson, H., Jakupsstovu, S.H.I., Røttingen, I., Belikov, S., Asthorsson, O., Blindheim, J., Jonsson, J., Krysov, A., Malmberg, S.A., Sveinbjornsson, S., 1998. Distribution, migration and abundance of Norwegian spring spawning herring in relation to the temperature and zooplankton biomass in the Norwegian Sea as recorded by coordinated surveys in spring and summer 1996. Sarsia 83, 117-127.

Monstad, T., Bjelland, O., Bergstad, O.A., 1998. Fish stocks and trophic ecology in the Norwegian Sea. Working Document to The Northern Pelagic and Blue Whiting Fisheries Working group, Copenhagen, 26 April-6 May. ICES C.M./ACFM 18, 29.

Nøttestad, L., Giske, J., Holst, J.C., Huse, G., 1997. A length based hypothesis to explain feeding migration in pelagic fish. Canadian Journal of Fisheries and Aquatic Science 56, 26-34.

Nøttestad, L., Olsen, E., 2004. Whales and seals: Top predators in the ecosystem., in: Skjoldal, H.R. (ed.), The Norwegian Sea Ecosystem. Tapir academic press, Trondheim, Norway, pp. 395-


80

434. Okunishi, T., Yamanaka, Y., Ito, S., 2009. A simulation model for Japanese sardine (Sardinops

melanostictus) migrations in the western North Pacific. Ecol. Model. 220, 462-479. Olsen, E.M., Melle, W., Kaartvedt, S., Holst, J.C., Mork, K.A., 2007. Spatially structured interactions

between a migratory pelagic predator, the Norwegian spring-spawning herring Clupea harengus L., and its zooplankton prey. J. Fish Biol. 70, 799-815.

Pitcher, T.J., 1986. Behaviour of teleost fishes. Chapman and Hall, London 715 pp. Prokopchuk, I., Sentyabov, E., 2006. Diets of herring, mackerel, and blue whiting in the Norwegian

Sea in relation to Calanus finmarchicus distribution and temperature conditions. ICES J. Mar. Sci. 63, 117-127.

Railsback, S.F., Lamberson, R.H., Harvey, B.C., Duffy, W.E., 1999. Movement rules for individual-based models of stream fish. Ecological Modelling 123, 73-89.

Rey, F., 2004. Phytoplankton: The grass of the Sea, in: Skjoldal, H.R. (ed.), The Norwegian Sea ecosystem. Tapir Academic Press, Trondheim, Norway, pp. 97-137.

Rose, K.A., Allen, J.I., Artioli, Y., Barange, M., Blackford, J., Carlotti, F., Cropp, R., Daewel, U., Edwards, K., Flynn, K., Hill, S.L., 2010. End-To-End Models for the Analysis of Marine Ecosystems: Challenges, Issues, and Next Steps. Marine and Coastal Fisheries: Dynamics, Managment, and Ecosystem Science. 2, 115-130.

Skjoldal, H.R., Dalpadado, P., Dommasnes, A., 2004a. Food webs and trophic interactions, in: Skjoldal, H.R. (ed.), The Norwegian Sea ecosystem. Tapir Academic Press, Bergen, pp. 447-506.

Authro, 1993a. Mare Cognitum Science Plan for Research on the Nordic Seas. Authro, 1993b. Mare Cognitum. Science Plan for Research Marine Ecology of the Nordic Seas

(Greenland, Norwegian, Iceland Seas) 1993-2000. Skjoldal, H.R., Sætre, R., Fernö, A., Misund, O.A., Røttingen, I., 2004b. The Norwegian Sea

Ecosystem. Tapir academic press, Trondheim 559 pp. Skjoldal, H.R., Sætre, R., Fernö, A., Misund, O.A., Røttingen, I. (eds.) 2004c. The Norwegian Sea

Ecosystem. Tapir, Trondheim 559 pp. Slotte, A., 1999. Differential utilization of energy during wintering and spawning migration in

Norwegian spring-spawning herring. J. Fish Biol. 54, 338-355. Sundby, S., 2000. Recruitment of Atlantic cod stocks in relation to temperature and advection of

copepod populations. Sarsia 85, 277-298. Travers, M., Shin, Y.J., Jennings, S., Cury, P., 2007. Towards end-to-end models for investigating the

effects of climate and fishing in marine ecosystems. Prog. Oceanogr. 75, 751-770. Utne, K.R., Hjøllo, S.S., Huse, G., Skogen, M., 2012. Estimating consumption of Calanus finmarchicus

by planktivorous fish in the Norwegian Sea estimated fromusing a fully coupled 3D model system. Mar. Biol. Res. 8, 527-547.

Utne, K.R., Huse, G., 2012. Estimating the horizontal and temporal overlap of pelagic fish distribution in the Norwegian Sea using individual-based modelling. Mar. Biol. Res. 8, 548-567.

Utne, K.R., Huse, G., submitted. Estimating overlap between pelagic fish in the Norwegian Sea by individual based modelling. Marine Biology Research.

Utne, K.R., Huse, G., Ottersen, G., Holst, J.C., Zabavnikov, V., Jackobsen, J.A., Óskarsson, G., Nøttestad, L., submitted. Horizontal distribution and overlap of planktivorous fish stocks in the Norwegian Sea during summers 1995-2006. Marine Biology Research.

Varpe, O., Fiksen, O., Slotte, A., 2005. Meta-ecosystems and biological energy transport from ocean


81

to coast: the ecological importance of herring migration. Oecologia 146, 443-451. Ware, D.M., Thomson, R.E., 2005. Bottom-up ecosystem trophic dynamics determine fish

production in the northeast Pacific. Science 308, 1280-1284. Weihs, D., 1973. Hydromechanics of Fish Schooling. Nature 241, 290-291. Zilanov, V.K., 1968. Some data on the biology of Micromesistius Poutassou (Risso) in the north-east

Atlantic. Rapports et Procès-Verbaux des Réunions du Conseil International pour l'Exploration de la Mer 158, 116-122.


82

9 On growth variations in Northeast Atlantic blue whiting and

resulting impacts on consumption

Verena Trenkel

Institut français de recherche pour l'exploitation de la mer (IFREMER), rue de l'île d'Yeu, BP 21105, 44311 Nantes cedex 3, France [email protected] Abstract Large variations in stock size have been observed in recent years for the blue whiting (Micromesistius poutassou) stock in the Northeast Atlantic. Growth variations leading to changes in survival are hypothesized to be a contributing factor. Using a modified von Bertalanffy growth model we examined the impact of trophic conditions and sea surface temperature on individual size variability (2004-2011) and mean weight-at-age (1981-2012). Due to sexual dimorphism males and females were analysed separately. The strength of the northeast Atlantic subpolar gyre was taken as an index for trophic conditions and recruit numbers to account for density-dependence. The results provide strong evidence for growth variations due to trophic conditions (mainly negative density-dependence) and a temperature effect (positive or negative depending on data set). As a consequence individual annual consumption varied in time, though this did not reduce much the estimated total stock consumption in years of high stock abundance. Keywords: von Bertalanffy, environmental effect, density-dependence. Introduction In the Northeast Atlantic massive changes in both abundance and distribution of the three largest pelagic stocks Norwegian spring spawning herring (Clupea harengus), Atlantic mackerel (Scomber scombrus) and blue whiting (Micromesistius poutassou) have occurred over the last two decades (Trenkel et al. 2014). The recruitment to the blue whiting stock experienced a four-fold increase during the period 1996-2005 (ICES 2011). After this period of high recruitment the situation has now returned to the low levels seen before 1996 (ICES 2013). Several linked hypotheses have been put forward to explain these large recruitment changes in terms of processes leading to changes in survival rates from eggs to recruits: larval dispersal processes, growth, prey availability, predation and stock structure (Payne et al. 2012). Based on a review of the literature, Payne et al. (2012) concluded that the two most likely hypotheses were i) changes in predation pressure by adult mackerel on blue whiting egg and larvae, and ii) prey availability. However, few blue whiting eggs and larvae have been found in mackerel stomachs (Pinnegar et al. 2014). In contrast, predation on young blue whiting is well documented. For example, in the Celtic Sea blue whiting up to a length of around 25 cm are consumed by a range of predators including hake, saithe and whiting (Pinnegar et al. 2003), in particular during feeding in the summer months (Trenkel et al. 2005). The mackerel predation hypothesis is not considered any further here. Prey availability for all life stages but in particular for larvae and juveniles is expected to affect survival directly and indirectly via growth as faster growing individuals move faster through the size window where they are most vulnerable to predation (Houde 1997). The diet of larval blue whiting


83

consists predominately of eggs, nauplii and copepodites of Calanus spp., Pseudocalanus spp., Arcatia spp. and Oithona spp., with little or no phytoplankton (Conway 1980, Hillgruber and Kloppmann 1999, Gonzalez-Quiros and Anadon 2001). Older blue whiting feed on copepods, euphausiids and amphipods in geographically varying proportions (Pinnegar et al. 2014). Blue whiting larvae are found along the shelf edge from the southern end of the Celtic Sea (about 48°N) to the north of Scotland (Bailey and Heath 2001, Ibaibarriaga et al. 2007). Their spatial distribution is shaped by the general oceanic circulation pattern (Hillgruber and Kloppmann 1999, Kloppmann et al. 2001) whereas their vertical distribution follows their diel feeding activity (Hillgruber and Kloppmann 2000). Larval growth rates vary across the distribution range (Bailey and Heath 2001, Brophy and King 2007), concomitant with temperature and salinity gradients (Bailey and Heath 2001). Phenotypic variability of somatic growth due to trophic conditions has been described for several species, including pelagic species such as herring (Shin and Rochet 1998) and capelin (Obradovich et al. 2014). Trophic conditions can be modified by changes in prey production, i.e. density-independent factors, but also by the number of competitors, i.e. density-dependence. The later leads to a year class effect in growth if it was caused primarily by intra-specific food competition during early life (pre-recruit) stages which have different diet requirements from older individuals. However, in addition to pre-recruits, density-dependent growth has also been demonstrated for recruited adult fishes (Lorenzen and Engberg 2002), including for cod, haddock and whiting in the North Sea (Bromley 1989, Bolle et al. 2004). Large scale bio-geographical shifts along the Northeast Atlantic shelf edge affecting several trophic levels from phytoplankton to blue whiting to pilot whales have been linked to general circulation patterns affecting hydrographical conditions (Hátún et al. 2009). These changes are linked to the strength of the subpolar gyre which is situated in the mid-Atlantic south of Greenland. In years with a weak subpolar gyre, warm, saline southern waters spread further to the north, covering Porcupine bank and up to the area south of Iceland (Hátún et al. 2009). In years of strong gyre, these areas are covered by colder, less saline waters. Hátún et al. (2009) provide evidence that the zooplankton abundance and composition varies with the strength of the subpolar gyre which they explained by advection effects rather than increased productivity. Higher abundance of more warm water zooplankton species but also higher phytoplankton densities occurred in years with weaker subpolar gyre. The strength of the subpolar gyre is described by the so called gyre index which is derived from the sea surface height field in the northern North Atlantic using multivariate analysis (Häkkinen and Rhines 2004). Alternatively or in addition to trophic induced phenotypic growth variations, temperature and oxygen can also affect growth (e.g. (Pörtner and Knust 2007)). Brunel and Dickey-Collas (2010) found a negative relationship between mean experienced temperature and asymptotic weight across geographically distributed herring stocks and a positive one for the growth rate k. Baudron et al. (2014) interpreted multi-decadal variations in cohort specific asymptotic length estimates of several species in the North Sea as being caused by the increase in bottom water temperature, but they did not investigate alternative explanations such food availability or density-dependence. To shed light on the importance and the factors affecting blue whiting phenotypic growth rate variations the following questions were studied in this paper: i) how much phenotypic variability exists in blue whiting growth? ii) are these variations density-independent, density-dependent or both? iii) is there evidence for a temperature effect on asymptotic growth? iv) how much has individual and total consumption varied over time? To address these questions growth models were fitted to individual length-at-age and mean


84

weight-at-age data for individuals aged 1 year and older and annual consumption was estimated. It was assumed that differences in larval growth, or more generally during the early months of life will not be completely compensated later in life but rather remain detectable in older fish. This assumption is supported by the observation that in years of high abundance young-of the-year blue whiting occupy a larger area on the feeding grounds on the Bay of Biscay shelf (Persohn et al. 2009) and are more abundant in the most northern feeding grounds, the Barents Sea (Heino et al. 2008), which indicates an overspill effect which could reduce intra-specific competition in young-of-the-year fishes, though it could also lead to individuals moving into less rich habitat as has been observed for herring (Reum et al. 2013). In contrast, for temperature, mean conditions over the life span of a fish were investigated rather than solely during the juvenile phase. Material and Methods Data sources Three data sets on growth were available for this study, individual length and age data from the international blue whiting survey for the period 2004-2011, individual data collected in 2011 onboard a commercial fishing vessel and mean weight-at-age in the catch (both sexes combined) calculated by the ICES stock assessment working group for 1981-2012 (ICES 2013). All data sets have shortcomings: the first and second ones are short, the second and third come from the catch, and the third one is average numbers and has both sexes combined. Therefor all data sets were used and the first two combined. Mean weight-at-age in the catch were available for ages 1 to 10+ years (ICES 2013). The plus group was removed from the analysis. The international blue whiting spawning stock survey taking place in March-April has been coordinated by ICES since 2004 (ICES 2011). Its main objective is to provide a spawning stock biomass estimate for stock assessment. To interpret the collected acoustic-backscatter information, pelagic trawl sampling is carried out opportunistically and from each haul otoliths of 30-100 individuals are randomly collected for age reading. Further, information on length (lower mm or 0.5 cm), weight (g), sex and maturity stage as well as the geographic position are stored. The biological data were available for the period 2004 to 2011 for individuals spanning a latitudinal range from 50.23 to 60.20°N. To ensure satisfactory latitudinal coverage and at least three years of sampling for all year classes, the data were restricted to latitudes ranging from 53° to 60.5°N (Figure 1a). The majority of individuals were 1 to 6 years old but individuals up to age 11 occurred. To balance the data as much as possible, only year classes 1999 to 2007 were considered. As age reading becomes more unreliable for older individuals, individuals of 10 years and older were removed. Combining this data with the data collected onboard the commercial fishing vessel slightly before the survey in January to March 20111, a length data set comprising 7983 females and 8090 males collected during 2004 to 2011 was obtained (Figure 1b).


85

2 4 6 8

10

15

20

25

30

35

40

AgeB

lue w

hitin

g length

(cm

)

Females

Males

(b)

Figure 1. (a) Map with selected haul locations for individuals sampled during 2004-2011. (b) Length-at-age of selected individuals by sex. In (a) H Hebrides terrace, R Rockall Bank, P Porcupine Bank. In (b) symbol radius is proportional to the number of observations and the ages were shifted to show males and females separately. A satellite altimetry derived gyre index for 1994-2013 and a simulation derived gyre index for the period 1960-2003 were supplied by H. Hátún (Faroes Island Marine Institute). Here the gyre index is used as a proxy for blue whiting larvae feeding conditions assuming a weak gyre, corresponding to positive gyre index values, provides better feeding conditions compared to a strong gyre (negative gyre index values). Estimated blue whiting recruit numbers (age 1) for 1981-2012 were taken from the final stock assessment run carried out by ICES (2013). They were estimated as a random walk independent of the spawning stock, i.e. without a stock-recruit relationship. Estimated stock numbers at age were taken from the same assessment run. Monthly sea surface temperature (SST) data for the period 1984 to 2011 were extracted from the HADISST data set for the main feeding area in the Norwegian Sea and around the Faeroe islands (ICES areas IIa and Vb). The spatial SST values (1°x 1°) were averaged by year. For each individual fish (or age group of each cohort), these mean annual SST values were averaged over its life time, from the year of birth to the year of capture as a proxy for the temperature it might have experienced. Model The generalised von Bertalanffy somatic growth function for body mass is, e.g. Essington et al. (2001) (see electronic supplementary material for model for body length)

dtadk

a eWW

1

1)0)(1(1 (1)

where Wa is weight at age a, W∞ the asymptotic weight, k the growth rate, t0 weight at a=0 and d describes the allometric scaling of consumption with body size; it is often assumed that d=2/3. The asymptotic weight depends on k, d and the rate of synthesis per unit of physiological surface H

d

k

HW

1

1

(2)


86

Shin and Rochet (1998) adapted the von Bertalanffy growth model to incorporate the relationship between trophic conditions and phenotypic plasticity of fish somatic growth. They showed that the amount of food Bt available for growth in year t will only affect asymptotic length L∞ via its impact on Ht

t

tt

N

BeH (3)

with eƐ the energy conversion efficiency and Nt the number of individuals sharing the resource. Two factors in turn affect the amount of food available, overall prey production Pt and the proportion of it which is consumed, determined by the prey predation rate pt and prey mortality m0 due to other predators and factors

t

t

tt P

mp

pB

0 (4)

To account for the diet of blue whiting larvae and juveniles and assuming that the variability in advection of zooplankton into the area is driven by the strength of the subpolar gyre as suggested by Hátún et al. (2009), we modelled prey production Pt as a linear function of the gyre index Gt in the year of birth tc

ctt GP (5)

where δ encapsulates the enrichment in prey independent of hydrological conditions, i.e. gyre status. We retained the Shin and Rochet formulation for describing the prey predation rate pt as being proportional to the number of competitors Ntc

, here the number of individuals born in the

same year tc (year class c)

ctt Np (6)

Putting the elements together our modified von Bertalanffy models for growth in year t accounting for trophic conditions in the year of birth tc=t-a are

)1/(10)(1(

)1/(1

3

)21

, 11

1 dtadK

d

at

at

ta eN

G

kW

(7)

where 0

1m

e ,

0

2m

e and

0

3m

. Year class specific asymptotic weight and length

are then )1/(1)1/(1

3

)21

1

1d

at

d

at

at

k

H

N

G

kW

at

(8)

We will refer to eq. (7) as weight-VBGF and to the equivalent in length as length-VBGF (see electronic supplementary material). The model assumes that asymptotic size (weight or length) is determined by trophic conditions in the first year of life rather than being a cumulative effect of annual trophic conditions over the course of the life up to the sampling date. To investigate a cumulative effect we used an additional model with year class specific W∞ (L∞) values but without any explanatory variables. Finally, temperature effects on growth were investigated by using mean life time SST instead of the gyre index in eq. 7. The non-linear model (eq 7) was fitted by least squares (nls function in R) setting the allometric scaling parameter to d=0.75 as recommended by Temming and Herrmann (2009). To test for significance of the density dependence (number of recruits) only, and/or a gyre or temperature effect, six models were compared using Akaike's information criterion (AIC): constant model (θ2=0,


87

θ3=0), density-dependent model (θ2=0), density-independent model (θ3=0, gyre or temperature) and both (recruit and gyre or temperature). Goodness-of-fit was assessed by calculating the squared Pearson correlation coefficient between predicted and observed values. Uncertainty intervals for the implicit W∞ (L∞) estimates from the best model (smallest AIC) were obtained by drawing 1000 random parameter values from a multivariate normal distribution with the mean parameter estimates and the estimated variance-covariance matrix, inserting these sampled parameter values into eq. (8) and taking the 2.5th and 97.5th percentile values. The year class specific W∞ (L∞) values were estimated as random effects using the nlme function in R. As blue whiting shows sexual dimorphism in growth (Quéro 1984); Fig. 1b), separate length-VBGFs were fitted for males and females. This could not be done for the weight-VBGF as mean weight-at-age was only available for both sexes combined. Note that the individual weight data from the survey were not useable as body weight varied strongly with maturity state. The effect of phenotypic grow variations on variation of total food consumption and the proportion consumed by recruits of the NE Atlantic blue whiting stock were investigated using the estimated year class values for H obtained from fitting the best weight-VBGS to the mean weight-at-age data. For this annual individual consumption ca,t for each age class was calculated as in Essington et al. (2001) and then multiplied by the estimated numbers-at-age in the stock Na,t to get an estimate of total consumption Ct

ta

a

ta

d

taat

a

tat cNA

WHNC ,

10

1

,

1

,10

1

,

(9)

where A is the assimilation efficiency and t-a refers to the year of birth, i.e. the year class. This approach is an approximation as it assumes numbers and weight-at-age are constant during the year while in reality numbers decrease and body weight increases. We set A=0.65 as in Essington et al. (2001). Results Trophic conditions in the year of birth influenced length and weight-at-age of blue whiting. For the length data set, the best model included a temperature and a density-dependent effect for both males and females (Table 1). However, for females the model with only a density-dependent effect had the same AIC value. The density-dependence and temperature model also had the smallest AIC for the combined sex mean weight-at-age data set (Table 1). Thus the effect of density-dependence on trophic conditions explained most of the variations in individual length and mean weight-at-age. In all cases the relationship with the number of recruits was positive (θ3>0, Table 2), which means that asymptotic size was lower for more abundant year classes. The relationship with mean life time SST was positive (θ2>0) for the short individual length data set and negative for the longer mean weight-at-age data set (θ2<0). The squared correlation between model predictions and observed length (weight) was > 0.71 for all models and data sets, indicating good model fits (Table 1). The best length-VBGF with both life time mean SST and density-dependent trophic conditions in the year of birth explained temporal variations in mean length-at-age for males and females rather well compared to models with constant L∞ (Figure 2top). Changes in trophic conditions and mean life time SST explained the observed increase in mean length in all age groups in the late 2000s. Table 1. Summary of model performance for blue whiting growth models: Akaike’s information criteria and squared correlation between fitted and observed values (in brackets). The mechanistic models with the smallest AIC are in bold.


88

Model Length-VBGF

Weight-VBGF

Females n=7983

Males n=8090

Combined n=179

Constant W∞ / L∞ 32557 (0.74) 30592 (0.71) -955 (0.91)

Density-dependence (θ2 = 0) 31870 (0.76) 29423 (0.74) -1077 (0.95)

Gyre (θ3 = 0) 32214 (0.75) 29743 (0.73) -1032 (0.94)

SST (θ3 = 0) 32560 (0.74) 30534 (0.71) -1078 (0.95)

Density-dependence & gyre 31870 (0.76) 29340 (0.75) -1077 (0.95)

Density-dependence & SST 31766 (0.77) 29045 (0.76) -1088 (0.96)

Cohort L∞ /W∞ (random effect) 31727 (0.77) 29101 (0.76) -1059 (0.97)

Although the best weight-VBGF model included SST, as the time series for SST was shorter, the model with only density-dependence fit to the longer weight-at-age time series is shown in Figure 3. This model explained a large part of the observed variations in mean weight-at-age, including the large drop in the early 2000s and the subsequent increase (Figure 3left).

Figure 2. Top: Blue whiting mean length-at-age (symbols) and fitted length-VBGF (continuous line)


89

allowing L∞ to vary by cohort as a function of recruit density and mean life time SST. The dashed horizontal lines are length-VBGF fits with constant L∞. Bottom: Model derived implicit L∞ values by year class (blue line, 95% uncertainty interval) and independent year class specific L∞ estimates (random effects model, symbols and line).

Figure 3 Left: Blue whiting mean weight-at-age (symbols) and fitted weighth-VBGF (continuous line) allowing W∞ to vary by cohort as a function of trophic conditions (density-dependent model). The dashed horizontal lines are weight-VBGF fits with constant W∞. Right: Model derived implicit W∞ values by year class (blue line, 95% uncertainty interval), constant W∞ (dashed horizontal line) and independent year class specific W∞ estimates (random effects model, symbols and line). To explore how well trophic conditions in the year of birth explained year class specific growth variations, VBGF models with distinct values of W∞ (L∞) for each year class were fitted as random effects. This model had the lowest AIC values only for female L∞ (Table 1). Overall, the time trends and magnitude of changes of year class specific agreed with those of the asymptotic size estimates derived from the best trophic length- and weight-VBGF (Figure 2bottom, Figure 3right). This general agreement is not surprising as the time trends in year class specific asymptotic size estimates were similar to the trend in recruit estimates over the whole study period (1981-2012), and for the general trend of the gyre index (Figure S2 & S3). For mean life time SST time trends were similar over the short period of the length data set (Figure S2). In contrast, for the whole study period cohort specific W∞ estimates tended to decrease in periods of increasing SST, implying a negative relationship between the signs of change of growth and mean life time SST (Figure S3); this is consistent with the negative sign for SST for the fitted model (Table 2). Year class specific asymptotic length varied between 34.7 and 38.2 cm (CV 3.1%) for females and between 30.4 and 33.7 cm (CV 3.6%) for males, with the variations of the two sexes being in phase although they were estimated independently. Year class specific asymptotic weight for both sexes combined varied from 248 to 313 g (CV 8.2%).


90

Table 2. Blue whiting growth model parameter estimates for models with density dependence (recruitment strength) and temperature effect for L∞ /W∞.

Parameter Estimate Std. Error 2.5% 97.5%

Males (length-VBGF)

K 0.432 0.008 0.415 0.449

t0 -2.101 0.062 -2.231 -1.978

θ1 -5.716 0.552 -6.813 -4.644

θ2 (SST) 1.053 0.059 0.939 1.172

θ3 (recruit.) 0.00123 0.00003 0.00117 0.00129

Females (length-VBGF)

K 0.353 0.007 0.339 0.368

t0 -2.356 0.070 -2.503 -2.218

θ1 -2.078 0.816 -3.676 -0.480

θ2 (SST) 0.846 0.083 0.684 1.010

θ3 (recruit.) 0.00103 0.00004 0.00096 0.0011

Combined sex (weight-VBGF)

k 0.681 0.063 0.557 0.808

t0 -5.036 0.468 -6.093 -4.201

θ1 0.731 0.079 0.582 0.896

θ2 (SST) -0.022 0.006 -0.035 -0.010

θ3 (recruit.) 6 10-7 2 10-7 3 10-7 1 10-6

Annual individual consumption by age class varied little (CVs 8.5-11.8%, Figure 4left). In contrast, total estimated consumption by blue whiting aged one year and older was around 3.8 million t in the 1990s, increased about three times up to 2003, followed by a decrease until 2010 and subsequent slight increase at the end (Figure 4right). Thus interannual variation in total consumption was rather variable (CV 47%). Estimated annual consumption amounted to 142-161% of blue whiting stock biomass. Assuming fixed or variable W∞ by year class did not have a large impact on total consumption estimates (mean difference: -1.5%; range: -3.9% to +0.5%). The results were more sensitive to the value used for the assimilation efficiency A (Figure 4right). The contribution of age 1 individuals to total consumption varied between 5 and 40% and was highest during the late 1990s at the time when stock biomass increased strongly (not shown).


91

Figure 4. Estimated mean individual and total annual consumption and stock size of blue whiting stock in NE Atlantic derived from best model with variable W∞. Discussion The results of this study indicate that interannual variability in blue whiting size and mean weight-at-age can be explained both by density-dependence (cohort strength) and experienced temperature. Persohn (2009) also found a negative cohort abundance effect on the size of juvenile (0-group) blue whiting in the Bay of Biscay, but no such effect in the North Sea or the gulf of Lions in the Mediterranean Sea. In the oligotrophic Mediterranean, feeding conditions could always be limiting, which would make it harder to detect density-dependence. Alternatively, the timing of the investigations might be important. Reum et al. (2013) found that for herring density-dependent limitations of growth only emerged late in the first year growing season. Blue whiting in the Bay of Biscay was investigated by Persohn at al. using data from the end of the growing season in autumn while the data for the gulf of Lions were collected in June. In this study both data sets reflect the situation after the first growing season. Individual blue whiting consumption varied by ± 20% between years of high and low growth. Intraspecific competitive interaction among individuals could have led to not enough food being available in years with larger cohorts. Diet information from years of high and low abundance would be needed to evaluate this hypothesis directly, but unfortunately data primarily exist for recent low abundance years (Keating et al. in prep). An alternative mechanism is linked to the ‘basin model’ under which individuals spread out into less suitable habitat when density increases (MacCall 1990); evidence for this has been found for Pacific herring (Reum et al. 2013). Spatial occupancy of demersal blue whiting juveniles in the Bay of Biscay and the Celtic Sea in autumn was found to be positively linked to abundance, while this seemed to be less the case for the more pelagic adults (Persohn et al. 2009). Similarly, blue whiting were found to spread northwards in the Norwegian Sea and into the Barents Sea in years of high abundance (Heino et al. 2008, Huse et al. 2012). To corroborate the habitat explanation it would however be necessary to demonstrate that the feeding conditions in these areas were poorer than in the core feeding areas. In addition to intraspecific competition, interspecific competition could also partially explain blue whiting growth variations. There is empirical and modelling evidence for interspecific competition between herring and blue whiting in the Norwegian Sea, with vertical displacement of herring into shallower areas in years of high blue whiting abundance and an increasing consumption of the Calanus finmarchicus production by both species and mackerel (Huse et al. 2012). Temperature had a negative impact on mean weight-at-age while the relationship was positive for the shorter time series with individual lengths. Baudron et al. (2014) also found a negative relationship between increasing temperature in the North Sea and decreasing asymptotic lengths of haddock, whiting, Norway pout and sole. Similarly, chub mackerel length at age 0 was negatively related to SST (Watanabe and Yatsu 2004). Thus these studies support our findings of a long term negative relationship between blue whiting growth changes and temperature. Also, although temperature affects growth directly via its effect on anabolism and catabolism, it can also act indirectly via prey availability so that apparent temperature relationships do not have to be directly causal. Several methodological issues might impact somewhat the results of this study. Possibly the most important one could be a country effect in the interpretation of otolith rings for age determination (ICES 2011). To minimize the age reader effect only individuals up to age 9 were included in the study. Second, in part of the length data set length was measured to the lower 0.5 cm while it was


92

to the millimetre in the remainder. This is expected to increase variability though some bias cannot be excluded. Third, average SST around the Faroe islands and in the Norwegian Sea was used as an indicator of the experienced temperature. This can be considered indicative only as individuals are commonly found deeper and not all feed in these areas given the spatial extent of the population. Fourth, though the gyre index which encapsulates the general hydrographic conditions during the larval phase had less explanatory power than mean life time SST, the two factors might be confounded as temperature varies with the strength of the subpolar gyre. Finally, the stock assessment estimate for age 1 (January 1 the year following birth) was used as a proxy for year class strength which assumes that the stock assessment estimates were at least correct in relative terms. In conclusion, this study provides strong evidence for phenotypic growth variations due to trophic conditions (mainly negative density-dependence) and a temperature effect (positive or negative depending on data set). As a consequence individual annual consumption varied in time, though this did not reduce much the estimated total stock consumption in years of high stock abundance. Acknowledgement We would like to thank Imares (the Netherlands), IMR (Norway), the Faroes Island Marine Institute and the Irish Marine Institute for making the survey data available. We are grateful to the crew of the pelagic trawler Joseph Roty 2 for collecting samples. This study received funding from the European Union Seventh Framework Programme project EURO-BASIN (ENV.2010.2.2.1-1) under grant agreement n° 264933 and a contract from FROM NORD to Ifremer. References Bailey M.C., Heath M.R., 2001, Spatial variability in the growth rate of blue whiting (Micromesistius

poutassou) larvae at the shelf edge west of the UK. Fisheries Research 50(1-2), 73-87. Baudron A.R., Needle C.L., Rijnsdorp A.D., Marshall C.T., 2014, Warming temperatures and smaller

body sizes: synchronous changes in growth of North Sea fishes. Global Change Biology 20(4), 1023-1031.

Bolle L.J., Rijnsdorp A.D., Van Neer W., Millner R.S., van Leeuwen P.I., Ervynck A., Ayers R., Ongenae E., 2004, Growth changes in plaice, cod, haddock and saithe in the North Sea: a comparison of (post-)medieval and present-day growth rates based on otolith measurements. Journal of Sea Research 51(3-4), 313-328.

Bromley P.J., 1989, Evidence for density-dependent growth in North Sea gadoids. Journal of Fish Biology 35, 117-123.

Brophy D., King P.A., 2007, Larval otolith growth histories show evidence of stock structure in Northeast Atlantic blue whiting (Micromesistius poutassou). ICES Journal of Marine Science 64, 1136-1144.

Brunel T., Dickey-Collas M., 2010, Effects of temperature and population density on von Bertalanffy growth parameters in Atlantic herring: a macro-ecological analysis. Marine Ecology Progress Series 405, 15-28.

Conway D.V.P., 1980, The food of larval blue whiting, Micromesistius poutassou (Risso), in the Rockall area. Journal of Fish Biology 16, 709–723.

Essington T.E., Kitchell J.F., Walters C.J., 2001, The von Bertalanffy growth function, bioenergetics, and the consumption rates of fish. Can. J. Fish. Aquat. Sci. 58(11), 2129-2138.

Gonzalez-Quiros R., Anadon R., 2001, Diet breadth variability in larval blue whiting as a response to plankton size structure. Journal of Fish Biology 59(5), 1111-1125.


93

Häkkinen S., Rhines P.B., 2004, Decline of subpolar North Atlantic circulation during the 1990s. Science 304(5670), 555-559.

Hátún H., Payne M.R., Beaugrand G., Reid P.C., Sandø A.B., Drange H., Hansen B., Jacobsen J.A., Bloch D., 2009, Large bio-geographical shifts in the north-eastern Atlantic Ocean: From the subploar gyre, via plankton, to blue whiting and pilot whales. Progress in Oceanography 80, 149-162.

Heino M., Engelhard G.H., Godø O.R., 2008, Migrations and hygrography determine the abundance fluctuations of blue whiting (Micromesistius poutassou) in the Barents Sea. Fisheries Oceanography 17, 153-163.

Hillgruber N., Kloppmann M., 1999, Distribution and feeding of blue whiting Micromesistius poutassou larvae in relation to different water masses in the Porcupine Bank area, west of Ireland. Marine Ecology Progress Series 187, 213-225.

Hillgruber N., Kloppmann M., 2000, Vertical distribution and feeding of larval blue whiting in turbulent waters above Porcupine Bank. Journal of Fish Biology 57, 1290-1311.

Houde E.D., 1997, Patterns and trends in larval-stage growth and mortality of teleost fish. Journal of Fish Biology 51 Suppl. A, 52-83.

Huse G., Holst J.C., Utne K., Nottestad L., Melle W., Slotte A., Ottersen G., Fenchel T., Uiblein F., 2012, Effects of interactions between fish populations on ecosystem dynamics in the Norwegian Sea - results of the INFERNO project Preface. Marine Biology Research 8(5-6), 415-419.

Ibaibarriaga L., Irigoien X., Santos M., Motos L., Fives J.M., Franco C., De Lanzos A.L., Acevedo S., Bernal M., Bez N., Eltink G., Farinha A., Hammer C., Iversen S.A., Milligan S.P., Reid D.G., 2007, Egg and larval distributions of seven fish species in north-east Atlantic waters. Fisheries Oceanography 16(3), 284-293.

ICES, 2011, Report of the Working Group on Northeast Atlantic Pelagic Ecosystems Surveys (WGNAPES), pp. 192.

ICES, 2011, Report of the working group on Widely Distributed Stocks (WGWIDE), pp. 642. ICES, 2013, Report of the Working Group on Widely Distributed Stocks (WGWIDE), pp. 944. Keating J., Minto C., Brophy D., Officer R., Pinnegar J.K., Trenkel V., in prep, The Dirichlet

multinomial distribution resolves overdispersed count data in blue whiting stomach content analysis.

Kloppmann M., Mohn C., Bartsch J., 2001, The distribution of blue whiting eggs and larvae on Porcupine Bank in relation to hydrography and currents. Fisheries Research 50(1-2), 89-109.

Lorenzen K., Engberg K., 2002, Density-dependent growth as a key mechanism in the regulation of fish populations: evidence from among-population comparisons. Proceedings of the Royal Society London, Series B 269, 49-54.

MacCall A.D., 1990, Dynamic geography of marine fish populations. Seattle, W.A., University of Washington Press.

Obradovich S.G., Carruthers E.H., Rose G.A., 2014, Bottom-up limits to Newfoundland capelin (Mallotus villosus) rebuilding: the euphausiid hypothesis. ICES Journal of Marine Science 71(4), 775-783.

Payne M.R., Egan A., Fässler S.M.M., Hátún H., Holst J.C., Jacobsen J.A., Slotte A., Loeng H., 2012, The rise and fall of the NE Atlantic blue whiting. Marine Biology Research 8(5-6), 475-487.

Persohn C., 2009, Sensibilité des populations de poissons aux changements globaux en fonction de leurs traits d’histoire de vie étudiée par une approche empirique, University of Brest, pp. 200.


94

Persohn C., Lorance P., Trenkel V.M., 2009, Habitat preferences of selected demersal fish species in the Bay of Biscay and Celtic Sea, North-East Atlantic. Fisheries Oceanography 18, 268-285.

Pinnegar J.K., Goñi N., Trenkel V.M., Arrizabalaga H., Melle W., Keating J., Óskarsson G., 2014, A new compilation of stomach content data for commercially-important pelagic fish species in the Northeast Atlantic. Earth Syst. Sci. Data Discuss. 7 197–223.

Pinnegar J.K., Trenkel V.M., Tidd A.N., Dawson W.A., du Buit M.-H., 2003, Does diet in Celtic Sea fishes reflet prey availability? Journal of Fish Biology 63 (Supplement A), 197-212.

Pörtner H.O., Knust R., 2007, Climate change affects marine fishes through the oxygen limitation of thermal tolerance. Science 315(5808), 95-97.

Quéro J.C., 1984, Les poissons de mer des pêches françaises. Paris, Grancher. Reum J.C.P., Essington T.E., Greene C.M., Rice C.A., Polte P., Fresh K.L., 2013, Biotic and abiotic

controls on body size during critical life history stages of a pelagic fish, Pacific herring (Clupea pallasii). Fisheries Oceanography 22(4), 324-336.

Shin Y.-J., Rochet M.-J., 1998, A model for the phenotypic plasticity of North sea herring growth in relation to trophic conditions. Aquatic Living Resources 11, 315-324.

Temming A., Herrmann J.P., 2009, A generic model to estimate food consumption: linking von Bertalanffy's growth model with Beverton and Holt's and Ivlev's concepts of net conversion efficiency. Can. J. Fish. Aquat. Sci. 66(4), 683-700.

Trenkel V.M., Huse G., MacKenzie B., Alvarez P., Arrizabalaga H., Castonguay M., Goñi N., Grégoire F., Hátún H., Jansen T., Jacobsen J.A., Lehodey P., Lutcavage M., Mariani P., Melvin G., Neilson J.D., Nøttestad L., Óskarsson G.J., Payne M., Richardson D., Senina I., Speirs D.C., 2014, Comparative ecology of widely-distributed pelagic fish species in the North Atlantic: implications for modelling climate and fisheries impacts. Progress in Oceanography http://dx.doi.org/10.1016/j.pocean.2014.04.030.

Trenkel V.M., Pinnegar J.K., Dawson W.A., du Buit M.H., Tidd A.N., 2005, Spatial and temporal structure of predator–prey relationships in the Celtic Sea fish community. Marine Ecology Progress Series 299, 257–268.

Watanabe C., Yatsu A., 2004, Effects of density-dependence and sea surface temperature on interannual variation in length-at-age of chub mackerel (Scomber japonicus) in the Kuroshio-Oyashio area during 1970-1997. Fish. Bull. 102(1), 196-206.


95

ANNEX Growth function in length

The von Bertalanffy growth function in terms of body size Lt is:

mtamK

a eLL

1

1)0)(1(1 (S.1)

md

bK

HaL

1

11

(S.2)

where m = b(d-1) +1 and a and b are the parameters of the weight-length relationship b

tt LaW .

The growth rate K in eqs. (S.1) and (S.2) is related to k in eq (1) as K=k/b.

)1/(10)(1(

)1/(1

3

)211

, 11

mtamK

m

at

atd

ta eN

G

bK

aL

(S.3)

)1/(11

)1/(1

3

)211

1

m

at

dm

at

atd

bK

Ha

N

G

bK

aL

at

(S.4) The parameters of the length-weight relationship were obtained by fitting the relationship only to immature, resting or spent individuals of both sexes in the survey data set to avoid that variable gonad weight during spawning influenced the parameter values (Figure S1).

15 20 25 30 35 40

0

50

100

150

200

250

300

350

Length (cm)

Weig

ht

(g)

W 0.0035 L3.1

Figure S1. Length-weight relationship for blue whiting (both sexes combined).


96

Figure S2. Normalised year class specific L∞ estimates (length data set) and normalised explanatory variables for trophic conditions. Note that the y-axis has been reversed for recruits to account for the inverse effect.

Figure S3. Normalised year class specific W∞ estimates (weight-at-age data set) and normalised explanatory variables for trophic conditions. Note that the y-axis has been reversed for recruits to account for the inverse effect.


97

10 Blue Whiting population modelling: progress, questions &

next steps

John K. Pinnegar1, Robert Thorpe1 and Verena M. Trenkel2 1Centre for Environment, Fisheries & Aquaculture Science (Cefas), Pakefield Road, Lowestoft, Suffolk, NR33 0HT, UK. [Tel. +44 (0) 1502 524229, Fax. +44 (0) 1502 513865, e-mail, [email protected]] 2 Institut français de recherche pour l'exploitation de la mer (IFREMER), rue de l’ile d’Yeu, BP 21105, 44311 Nantes cedex 3, France. 1. Introduction/Methods Blue whiting (Micromesistius poutassou) is a small pelagic gadoid that is widely distributed in the eastern part of the North Atlantic. The highest concentrations are found along the edge of the continental shelf in areas west of the British Isles and on the Rockall Bank plateau where it occurs in large schools at depths ranging between 300 and 600 meters. Adults reach maturation at 2 – 7 years old and undertake long annual migrations from the feeding grounds (in the north and south) to the spawning grounds off the Porcupine Bank and along the shelf edge. Most of the spawning takes place between March and April. Juveniles are abundant in many areas, with the main nursery area believed to be the Norwegian Sea. The blue whiting population supports important commercial fisheries (with a total catch of around 103,592 tonnes in 2011) and is assessed by ICES as a single stock, spanning the whole northeast Atlantic from the Bay of Biscay to the Barents Sea (including Iceland and Greenland). In recent years, several authors have suggested that the blue whiting population might in fact represent a number of distinct sub-stocks (Payne et al. 2012) and that assessment methods should be devised that can accommodate multiple geographic areas, with limited mixing between each. Within Euro-Basin Cefas and IFREMER have committed themselves to revisiting the blue-whiting sub-model of an existing 3 species ‘GADGET’ model, originally developed for the Celtic Sea under the FP5 project DST2 (ended in 2004). GADGET (see www.hafro.is/gadget/ also Begley and Howell 2004) is a framework that allows multi-area, multi-fleet, multi-stock and multi-species simulations. It is both length and age-based, thus allowing detailed stock assessment, even where age determination is problematic. The original Celtic Sea model was parameterised for cod Gadus morhua, whiting Merlangus merlangius and blue whiting in ICES areas VIIe-k. It was primarily used to investigate predator-prey interactions (Trenkel et al. 2004). Within Euro-basin however, the focus has been solely on blue whiting with updates to include recent survey and catch datasets (from ICES stock assessment reports), implementation of 3 areas (south, central, north) and implementation of growth or recruitment functions that can take account of climate variability.

The revised model has been developed for the time period 1984 – 2011, with quarterly time-steps. Growth is derived according to a von Bertalanffy growth function with 16 length classes (from 9-41cm). Ten age classes (1-10) are assumed, and M = 0.2 at all ages. Key data sources include


http://www.hafro.is/gadget/


98

landings data for each region and quarter, commercial and survey catches by length for each region, an age-length key (numbers by length and age) within survey and fishery catches and a survey index for each region (EVHOE-south, IBWSS-central, IESSNS-north). Figure 1 illustrates the spatial structure of the newly-derived GADGET model. The southern area (Area 1) comprises ICES sub-divisions 7e-k, VIIIa-c, IXa, the central area (Area 2) comprises ICES sub-divisions VIa,b, VIIb,c, and the northern area (Area 3) comprises ICES sub-divisions IIa, Va,b.

Figure 1. Spatial structure of the 3 area GADGET model for blue whiting in the northeast Atlantic. Very little information is available concerning the migration characteristics of blue whiting. There have been no tagging studies on this species and therefore assumptions have been based on the relative distribution of commercial catches at different times of year (as reported in the most recent ICES stock assessment). At present, the model assumes:

1. In spring the central fish population (area 2) produces recruits but no populations move

2. In summer 10% of the central population moves south (to area 1) and 80% north (to area 3), the existing northern and southern populations stay put

3. In autumn 5% of the northern and southern populations move to the central area, the rest stay put

4. In winter there is a 5% migration from the south and 20% from the north whilst the middle population stays put.

In Gadget there are 7 ‘off the shelf’ growth models that can be implemented, some of which take account of climatic factors or food availability. The blue whiting model currently uses ‘option 2,’ where the growth can be modified using temperature. A default value of 5°C is used throughout,


99

but it is intended that a time-series of sea-surface temperature (SST) will be introduced for the 3 regions and used to explore sensitivity. There are 4 ‘off the shelf’ recruitment models available within Gadget, however the approach used at present is to simply read in a file from 1981-2011 based on the ICES stock assessment. In the longer term (as part of Euro-basin) the intention will be incorporate a recruitment function that takes account of climate variability (either SST, NAO, AMO, salinity or some other climatic index such as a regional gyre index), and thereby try to replicate some of the features described by Hátún et al. (2009). These authors suggest that that the spawning distribution of blue whiting is variable, and regulated by the hydrography west of the British Isles. When the North Atlantic subpolar gyre is strong it spreads cold, fresh water masses east over Rockall Plateau, the spawning is constrained along the European continental slope and in a southerly position near Porcupine Bank. When the gyre is weak, conditions are relatively saline and warm, the spawning distribution moves northwards along the slope and especially westwards covering Rockall Plateau. The apparent link between the spawning distribution and the subpolar gyre is the first step towards understanding recruitment variability of the blue whiting stock.

Using the estimated number of individuals in each area, by quarter and length class the total quantity of food consumed has been calculated assuming ‘gastric-evacuation’ equations developed for cod feeding on krill (Temming & Herrmann 2003). At the moment we have assumed a flat 5°C SST in each area and season, and identical evacuation rates for all prey types. The longer-term aspiration under Euro-Basin will be to use locally relevant (and seasonally representative) temperatures to calculate feeding rate, as well as to explore whether ‘gastric-evacuation’ equations developed for mackerel (Temming et al. 2002) are more appropriate in this context, rather than those developed for cod. 2. Results Figure 2 shows that the 3-area Gadget model, provides a broadly consistent picture of the blue whiting stock with the most recent ICES stock assessment, at least in terms of temporal trajectory and age-structure. A key difference with the ICES assessment is that outputs (in terms of biomass or numbers) can be generated for each geographic area, for each season (quarter of the year) and for each length class, as well as for each age class. This information is particularly useful when trying to quantify the total amount of food consumed by blue whiting in the north-east Atlantic.

0

5E+11

1E+12

1.5E+12

2E+12

2.5E+12

Num

ber

of in

divi

dual

s

Year

Gadget (quarter 3) numbers-at-age

Age 10+

Age 9

Age 8

Age 7

Age 6

Age 5

Age 4

Age 3

Age 2

Age 1

0

20000

40000

60000

80000

100000

120000

140000

160000

Nu

mb

er

of

ind

ivid

ua

ls (

mil

lio

ns)

Year

ICES Stock Assessment numbers-at-age

Age 10+

Age 9

Age 8

Age 7

Age 6

Age 5

Age 4

Age 3

Age 2

Age 1

Figure 2. Number of individual blue whiting at age in the population, in quarter 3 of the year –


100

according to: (a) the 3-area Gadget model, and (b) the 2012 ICES stock assessment, based on a SAM model (Nielsen, 2009).

0

5E+11

1E+12

1.5E+12

2E+12

2.5E+12

Num

ber

of in

divi

dual

s

Year

Quarter 1 (spring)

Area 3

Area 2

Area 1

0

5E+11

1E+12

1.5E+12

2E+12

2.5E+12

Num

ber

of in

divi

dual

s

Year

Quarter 3 (autumn)

Area 3

Area 2

Area 1

Figure 3. Number of individual blue whiting in the population by geographic area – in quarter 1, (Jan, Feb, Mar) and quarter 3 (Jul, Aug, Sep). Figure 3 shows that the number of individuals present in each of the three geographic regions varies considerably throughout the year and in different years of the time series. In quarter 1 (the spawning season), the vast majority of blue whiting in the population are found in the central region (Area 2) with tiny numbers being retained in Area 3 (the north) and Area 1 (the south). By contrast, in quarter 3 (the feeding season) most blue whiting are found in the northernmost region (Area 3) with few individuals in the southern and central areas. When the numbers-at-length, by quarter are used to calculate the total amount of food consumed (aggregated across the whole year), it is possible to see that broadly equivalent quantities are consumed in area 2 and 3 but much less in area 1 (Figure 4). These results are still very preliminary, and do not take account of seasonal or geographic differences in consumption rate per capita (as a result of temperature etc), but they do give an overall picture of the enormous magnitudes involved. The DAPSTOM stomach content dataset for blue whiting (see Pinnegar et al., above) shows considerable variability in diet composition across the geographic range of this species with euphausids dominating in terms of numerical dominance in Iceland, the Bay of Biscay and the Irish Sea, but hyperiid amphipods dominating in the Norwegian Sea, eastern Greenland and the Celtic Sea. Efforts will be dedicated in the coming months, towards improving quantitative estimates of key zooplankton species consumed by blue whiting – and comparing these estimates with observed densities/production figures for euphausiids or hyperiids in the Euro-Basin area to assess the predation mortality imparted by planktivorous fish.


101

0

2E+14

4E+14

6E+14

8E+14

1E+15

1.2E+15

1.4E+15

1.6E+15

1.8E+15

2E+15

19

84

19

85

19

86

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

20

05

20

06

20

07

20

08

20

09

20

10

20

11

Food

Con

sum

ed (

tonn

es)

Year

Food consumed by blue-whiting in each area

Area 3

Area 2

Area 1

Figure 4. Total quantity (wet weight) of food consumed by blue whiting in each geographic area per year, estimated by the 3-area Gadget model. 3. Discussion The ICES stock assessment for blue whiting in the northeast Atlantic suggests that the 2005-2010 year classes were among the smallest ever observed. Spawning stock biomass (SSB) has declined from a historical peak in 2003-2004 of more than 7.1 million tonnes to just above 2.8 million at the beginning of 2011. These patterns are largely replicated in the 3-area Gadget model (Figure 1), although this is perhaps not surprising, given that a time-series of recruits is currently inputted directly into the Gadget model and this is itself derived from the ICES stock assessment. Discussion with Gadget experts has suggested that it is possible (and not too hard) to write a new recruitment relationship and implement it in the Gadget code, but this may not be needed. Any of the existing four recruitment functions currently available in Gadget can incorporate temperature through a "time variable". This simply replaces a single parameter to be estimated with a time series of parameters, that can incorporate arithmetic functions. This is the approach taken by Andonegi et al. (2011). The North Atlantic subpolar gyre has previously been shown to have a strong influence on the behaviour of the blue whiting stock stock (Hátún et al. 2009). Synchronous changes in the gyre and recruitment suggest a causal linkage and the possibility of forecasting recruitment. A range of mechanisms were reviewed by Payne et al. (2012) that may explain these observed changes, with two major candidate hypotheses being identified. One hypothesis suggests that the large mackerel (Scomber scombrus) stock in this region may feed on the pre-recruits of blue whiting, with the spatial overlap between blue whiting and mackerel being regulated by the subpolar gyre. Alternatively, variations in the physical environment may have given rise to changes in the amount, type and availability of food for larvae and juveniles, impacting their growth and survival and therefore recruitment (Payne et al. 2012). In the present study we have assumed that a single blue whiting stock exists, with mixing in area 2 (the ‘central’ spawning area) in spring. Genetic evidence also suggests that the vast majority of blue whiting belong to a single Hebrido-Norwegian stock (Mork & Giaever 1995; Varne & Mork 2004; Ryan et al. 2005), although a genetically distinct population may exist in the Barents Sea. This


102

Barents Sea stock is comparatively small compared to the Hebrido-Norwegian stock. It has been suggested to spawn on the northern Norwegian continental slope (Zilanov 1968) and could be modelled as a separate geographic region within Gadget model in the future if desired. References Andonegi, E., Fernandes, J. A., Quincoces, I., Irigoien, X., Uriarte, A., Pérez, A., Howell, D., and

Stefánsson, G. 2011. The potential use of a Gadget model to predict stock responses to climate change in combination with Bayesian networks: the case of Bay of Biscay anchovy. – ICES Journal of Marine Science, 68: 1257–1269.

Begley, J., and Howell, D. 2004. An overview of Gadget, the globally applicable Area-Disaggregated General Ecosystem Toolbox. ICES Document CM 2004/FF: 13. 16 pp.

Háún, H., Payne, M.R., Jacobsen, J.A., (2009) The North Atlantic subpolar gyre regulates the spawning distribution of blue whiting (Micromesistius poutassou). Can. J. Fish. Aquat. Sci. 66: 759–770.

Mork J, Giaever M 1995 Genetic variation at isozyme loci in blue whiting from the north-east Atlantic. J Fish Biol 46: 462-468.

Nielsen, A. 2012. State-space models as an alternative to overparametrized stock assessment models. In preparation.

Payne, M.R., Egan, A., Fässler, S.M.M., Hátún, H., Holst, J.C., Jacobsen, J.A., Slotte, J., Loeng, H. (2012) The rise and fall of the NE Atlantic blue whiting (Micromesistius poutassou). Marine Biology Research, 8: 475-487.

Ryan, A. W., Mattiangeli, V., and Mork, J. 2005. Genetic differentiation of blue whiting (Micromesistius poutassou Risso) populations at the extremes of the species range and at the Hebridese-Porcupine Bank spawning grounds. ICES Journal of Marine Science, 62: 948-955.

Temming A, Herrmann JP (2003) Gastric evacuation in cod: prey-specific evacuation rates for use in North Sea, Baltic Sea and Barents Sea multi-species models. Fish Res 63: 21−41.

Temming A, Bøhle, B., Skagen, D.W., Knudsen, F.R. (2002) Gastric evacuation in mackerel: the effects of meal size, prey type and temperature. J Fish Biol, 61, 50–70

Trenkel, V.M., Pinnegar, J.K., Blanchard, J.L., Tidd, A.N. (2004) Can multispecies models be expected to provide better assessments for Celtic sea groundfish stocks? ICES CM 2004/FF:05

Varne, R. and Mork, J (2004) Blue whiting (Micromesistius poutassou) stock components in samples from the northern Norwegian Sea and Barents Sea, winter 2002. ICES CM 2004/EE:16

Zilanov, V.K. 1968. Occurrence of Micromestius poutassou (Risso) larvae in the Norwegian Sea in June 1961. Journal du Conseil Permanent International pour l’Exploration de la Mer, Rapports et Proces-Verbaux des Reunion 158: 116-122.

seventh framework programme theme 7 environment ...eurobasin.dtuaqua.dk/eurobasin/documents... ·...

Documents