© 2005 International Council for the Exploration of the Sea
Analysis of prey size preference of North Sea whiting, saithe, and grey gurnard
Zentrum für Meeres und Klimaforschung, Institut für Hydrobiologie und Fischereiwissenschaft Olbersweg 24, D22767 Hamburg, Germany
*Correspondence to J. Floeter: tel: +49 40 42838 6611; fax: +49 40 42838 6618. e-mail: jfloeter{at}uni-hamburg.de.
Size preference for prey fish of North Sea whiting, saithe, and grey gurnard was analysed. The analysis combined size-specific prey abundance estimates derived from bottom-trawl surveys with size frequencies of prey in predator stomachs from the International North Sea Stomach Database. To estimate the abundance of all potential prey fish in the sea, predator-specific length-based number spectra were calculated. Prey spectra were weighted by local predator abundance to take the spatial–temporal overlap between predator and their prey into consideration. Species-specific prey size preference models are presented. Contrary to former results, the preferred predator–prey weight ratio of whiting and grey gurnard is an exponentially increasing function of predator size and an exponentially decreasing function of the slope of the number spectrum. When predators grow, they prefer larger prey in absolute units. However, from a species-specific body size onwards they increasingly shift their prey preference towards relatively smaller prey sizes. From a bioenergetic point of view, this behaviour most likely maximizes the predator's foraging efficiency by reducing the expenditure of costly, anaerobically generated energy expended during burst swimming.
Keywords: bioenergetics, biomass spectrum, diet selection, gurnard, optimal foraging, prey size preference, saithe, whiting
Received 22 December 2003; accepted 18 March 2005.
 |
Introduction |
|---|
 Top Introduction Material and methodsResultsDiscussionReferences |
|---|
The determination of the relationships between predator and prey sizes that maximize the predator's net energy gain plays a central role in optimal foraging studies. In this analysis, we determine and parameterize the size preference function of North Sea whiting (Merlangius merlangus) and grey gurnard (Eutrigla gurnadus) that feed on fish; prey size preference of saithe (Pollachius virens) feeding on fish is described qualitatively.
Whiting is one of the most abundant and widely distributed gadoids in the North Sea. Crustaceans such as euphausiids and crangonid shrimps account for approximately 50% of the diet of small whiting (<15 cm; Daan, 1989; Hislop et al., 1997). The percentage of fish prey increases rapidly with predator size. Whiting larger than 30 cm feed almost entirely on fish, predominantly sandeels, Norway pout (Trisopterus esmarkii), and clupeids (Temming et al., 1998). In the northern and northeastern North Sea saithe is an important predator on sandeels, clupeids, Norway pout, and haddock (Melanogrammus aeglefinus; Daan, 1989; Hislop et al., 1997). Grey gurnard is also a widely distributed demersal species in the North Sea, and has been ranked in the 10 dominant species (Yang, 1982; Daan et al., 1990). An increasing contribution of fish prey to the diet corresponding to a decreasing contribution of crustacean prey with increasing predator size has been reported for all three species (Daan, 1989; Hislop et al., 1997). Such a shift in the diet appears to be a general pattern in many predatory fish species, although the size at which these shifts occur may be different (Temming et al., 1998). In grey gurnard, the size at which they switch prey is relatively small (20 cm). Larger grey gurnard feed primarily on juvenile fish, a large proportion of which consists of commercially exploited species, including whiting and sandeel (de Gee and Kikkert, 1993). The magnitude of annual fish biomass consumed by grey gurnard is comparable to that of whiting (ICES, 2002).
The main difficulty in the evaluation of selective feeding at the individual level is to estimate the prey abundance in the ambient environment at the scale of the predator's true perception. The reconstruction of the actual feeding situation in the North Sea at a spatial and temporal scale small enough to be relevant to individual predators is not possible from the integrated information obtained from standard trawl surveys. Instead of following an individual-based mechanistic approach, a macroscale model in terms of a size spectrum of potential fish prey in the North Sea was developed in a recent study of the diet composition of North Sea cod (Gadus morhua, Floeter and Temming, 2003). For a review of the methods needed to differentiate between preference, electivity, acceptability, and apparent preference, see Underwood et al. (2004).
The goal of the present analysis was to estimate the average predator-specific prey fields, i.e. the abundance and size structure of potential fish prey at the ecosystem level, and to identify the average selectivity of North Sea whiting, saithe, and grey gurnard with respect to the size of prey fish. Additionally, predator-specific prey size preference models are presented. Finally, it is discussed why the predator-specific prey size preferences lead to a maximization of foraging efficiency by limiting their expenditure of costly, anaerobically generated energy.
|
Material and methods |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Stomach content data
Data on the diet of the three predator species were obtained from the International North Sea Stomach Database (INSSB; Daan, 1989; Hislop et al., 1997). Samples had been taken under the auspices of the International Council for the Exploration of the Sea (ICES) in each quarter of the year, and the spatial resolution was the ICES statistical rectangle (30 nm x 30 nm). Detailed descriptions of sampling and stomach content analysis procedures can be found in Robb (1991). The sampling periods and the number of stomachs sampled varied between species (Tables 1,2,3).
|
|
|
Only fish prey for which data on total length were available were used for this study. The total length of fish prey was converted to weight assuming isometric growth. Predator weight was calculated from empirical length–weight relationships (Coull et al., 1989). Owing to the limited amount of stomach content data in the extreme size classes, a year-specific analysis was not possible for whiting and saithe, and data of all years were pooled by quarter. Quarterly grey gurnard samples were taken only in 1991. During the sampling period, different predator and prey size classes were used during the surveys (Hislop et al., 1997). Therefore, the data had to be re-categorized to maximize the usable amount of information while keeping the size classes small (Table 4). Predator size classes were only included in the analysis when a minimum of 50 full stomachs had been sampled in a quarter. Predator–prey size class interactions were only included in the analysis when a minimum of 10 prey items had been recorded.
|
Fish abundance data
To estimate the size-structured abundance of prey fish in the field, data from 12 years (1980–1991) of the first quarter International North Sea Young Fish Survey (IYFS) and 5 years (1991–1995) of the second and third quarter International North Sea Bottom Trawl Survey (IBTS) were used. These standardized surveys covered the entire North Sea, and a Grande Ouverture Verticale (GOV) bottom trawl was predominantly used for sampling (ICES, 1992).
Combination of stomach and abundance data
Stomach samples of the first quarter were predominantly taken from the standard hauls of the IYFS. Therefore, most of the first quarter stomach and abundance data used in this analysis were gathered from the same trawl tows. Stomach and prey abundance data from the second and third quarters originate from different trawl hauls, because field abundance data from the second and third quarters of the period 1980–1990 were not available, and estimation of fish prey abundance was based on IYFS data from 1991 to 1995 instead. The second and third quarter stomach content data stem from the period 1980 to 1991.
Estimation of fish prey in the sea by a number spectrum
Data on prey fish abundance obtained from the IYFS and IBTS were used for the calculation of the number spectra. To take the spatial and temporal overlap between the predators and their prey into consideration, the abundance of prey fish was weighted at the ICES statistical rectangle scale with the catch per unit effort (cpue, i.e. numbers caught per hour of trawling) of the predator species. To reduce the bias introduced by species-specific gear catchability, the field data on prey fish abundance were corrected using a modified version of a method developed by Yang (1982). Eventually, size-based number spectra (Pope and Knights, 1982) of the fish prey size distribution in the field were calculated for the unbiased size range and extrapolated towards smaller sizes (Table 5). A detailed description of the calculation procedure is provided in Floeter and Temming (2003).
|
Index of preference
We used the index of relative electivity (Ei*) and the selectivity coefficient (Wi) (Vanderploeg and Scavia, 1979) as a mathematical index of preference. Both are based on Chesson's (1978)
and provided equal results in the present analysis. As the index of relative electivity takes the number of prey types into account it has the convenient property of indicating random-feeding with a zero-value within a range from –1 to 1. Thus, the index Ei* is used for graphical presentation and Wi for fitting the size preference model, i.e. the peaks of the size preference functions based upon the index Wi were used for the calculation of the preferred predator–prey weight ratios.
Preference function
The stomach content data for all available years and quarters were pooled. The general prey size preference function (Equation (1)) developed for North Sea cod (Floeter and Temming, 2003) was fitted to the selectivity coefficients (Wi) obtained for whiting, saithe, and grey gurnard, using the Levenberg–Marquardt algorithm for non-linear regression in SPSSTM 10.
|
| (1) |
|
Results |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
The relative preference of whiting for fish prey of different sizes, expressed as the index of relative electivity (Ei*) shows that the smallest prey size class was not selected (depicted by a negative Ei*) in any quarter of the year (Figure 1). The selectivity for prey size increased with an increase in predator length. For almost all predator sizes the curve is restricted to the increasing left side of the function. Only in the first and fourth quarters do the predator size classes 27.5 cm and 35 cm display a small peak or plateau, i.e. the preference differences between neighbouring prey size classes become smaller. For the fourth quarter, predators appear to prefer smaller prey than in the remaining seasons. The preference curves of mid-sized whiting are generally similar.
|
Saithe also showed an increasing preference for larger prey with an increase in predator length. For all predator sizes the curve is restricted to the increasing left part of the function (Figure 2). With the exception of the first quarter, the differences between different predator size classes were more pronounced than those in whiting.
|
For grey gurnard too few data were available for the first and fourth quarters, so no reliable size preferences could be determined for those periods. In the other two quarters, grey gurnard appeared to show an ontogenetic increase of preferred prey size, although the absolute prey sizes are smaller than those for whiting and saithe (Figure 3). The peaks of the preference functions were poorly defined. Only in the third quarter does the second largest predator size class show a decrease in preference for the largest prey sizes.
|
When taking the temporal coverage of the data into account (Tables 1,2,3), the quarterly differences between preference curves were generally small for all three predator species. The most notable differences in preferred absolute prey size occurred in small whiting (Figure 4). Saithe of twice the size of whiting preferred the same absolute prey sizes. Small grey gurnard preferred approximately equal prey sizes as equally sized whiting predators, whereas large grey gurnard preferred almost half the size of fish large whiting preferred to prey upon.
|
The predator–prey weight ratios (PWRs) of whiting showed a clear increase with predator size for mid-size to large whiting (Figure 5). Smaller whiting showed a tendency for an initial decrease in PWR with increasing predator size. However, the magnitude of decrease differed between quarters. With the exception of the first quarter, the PWR of saithe decreased as a function of predator size. Only saithe larger than 80 cm switched to a preference for relatively small prey. The PWR function in the first quarter resembled the function for whiting, with an initial decrease followed by a steady increase. In quarter 2 and particularly quarter 3, grey gurnard showed a steadily increasing PWR with an increase in predator size.
|
The species comparison revealed that saithe has by far the highest PWRs, i.e. on the relative scale saithe preferred the smallest prey (Figure 6). The most obvious PWRs in the third quarter (Figure 6c) showed that the initial PWRs for small whiting and grey gurnard were at similar levels. However, with increasing predator size, the PWR of grey gurnard increased at a much higher rate. Thus, large whiting preferred the relatively largest prey.
|
The general prey size preference function (Equation (1)) for whiting explained 97% of the variance with 494 degrees of freedom (Table 6). The general prey size preference function for grey gurnard explained 96% of the variance with 275 degrees of freedom (Table 7). Following Floeter and Temming (2003), the models incorporate the effect of predator size as a linear function and the effect of changing slopes on the log of preferred predator–prey weight ratio as a power function. For saithe, the resulting quarterly preference functions were not unambiguous, so no model was fitted to the data. The confidence intervals for the parameter estimates in the whiting model were much narrower, although the grey gurnard model was fitted only to the results from the data-rich second and third quarters.
|
|
|
Discussion |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Methodology
In order to achieve a more detailed analysis of prey size preference, it would have been desirable to analyse the data year- and quarter-specific data. However, data availability did not allow this approach. Thus, we aggregated the stomach data from different years by quarter. This method provided general prey size preference curves, which probably cannot be applied to small-scale feeding situations, where individual predators tend to cluster-sample their prey from the locally available prey fields. On the other hand, the aggregation of information may have stabilized the resulting preference functions. To enhance the credibility of the results, the local prey weighting procedure ensures that prey availability at the intermediate scale, i.e. between the individual scale and the ecosystem level, is taken into account. Therefore, the results should be valid for the upper mesoscale to a North Sea ecosystem level.
The methodology applied in this analysis has been successfully developed and applied in a recent explanatory analysis of the diet composition of North Sea cod (Floeter and Temming, 2003). A sensitivity analysis performed in the analysis of North Sea cod demonstrated that the methods are robust with regard to possible over- or underestimation of available prey abundances in the field. Results of this sensitivity analysis showed that the trend of increasing PWR of North Sea cod is likely to be even more pronounced than the reverse. It can be assumed that these results are also applicable to the current analysis of other species' size preferences.
Prey size preference
This analysis confirms results from earlier studies for a greater variety of species, which indicated that the preferred predator–prey weight ratio is not independent of the predator size (Bundgaard and Sparholt, 1992; Floeter and Temming, 2003). The results disagree with the traditional view of constant preferred predator–prey weight ratios (Ursin, 1973). The preferred predator–prey weight ratio of North Sea cod feeding on fish is an exponentially increasing function of predator size and an exponentially decreasing function of the slope of the number spectrum (Floeter and Temming, 2003). This result is confirmed for North Sea whiting and grey gurnard. For both species, this study is the first size-based analysis of prey size selection. A former analysis of prey size preference of whiting (Ursin and Arntz, 1985) and grey gurnard (Ursin, 1975) followed Ursin's (1973) approach, which is restricted to constant preferred predator–prey weight ratios. Unfortunately, owing to the contradicting quarterly functions, the final answer for a general size preference function of North Sea saithe feeding on fish cannot be provided from this analysis. To achieve this, further field sampling would be necessary.
In their analysis of Baltic Sea whiting diet selection, Ursin and Arntz (1985) differentiated two feeding modes, preferential feeding, and size-indiscriminate feeding which included large numbers of invertebrate prey much smaller than the preferred body size. The latter feeding mode was hypothesized to occur in predators under food scarcity, as may be the case for intermediate size and large whiting in the Baltic Sea or for large cod in the North Sea (Daan, 1973). Following Ursin and Arntz (1985), the increase of PWR in larger predators observed in our study may be a result of non-selective feeding in a North Sea ecosystem, where the fishery has vastly reduced the numbers of large fish, rather than a change in size preference with predator size.
Bundgaard and Sparholt (1992) suggested an adaptive energy-saving behaviour as one reason for growing predators preying on relatively smaller prey: instead of wasting energy in searching for larger but scarce prey, predators would save energy when feeding on smaller, more abundant prey.
However, searching for prey is an aerobic activity. Another explanation for the observed increase of PWR in large predators, regardless of the species, can be drawn from more detailed bioenergetic analyses: Goolish (1991a) argued that the energy that can be gained on a sustainable basis for a hunting fish predator decreases sharply with increasing predator size as a consequence of the scaling of the energetic costs of activity. In detail:
- The relationship of the energy required for short-term burst swimming scales positively to body weight to the power of b
1.5, i.e. larger fish have higher weight-specific energy demands to overcome the drag forces (Webb, 1975; Somero and Childress, 1980).
- The highly efficient aerobic energy supply of muscle tissue scales negatively allometrically to body weight with the power of b 0.75, because of limits in oxygen uptake and blood circulation rates.
- Therefore, the activity threshold above which the energy has to be generated by the white muscle tissue on a less efficient anaerobic basis (the so-called anaerobic scope), is an increasing function of swimming speed and decreases with increasing bodyweight, i.e. scales negatively with body weight.
Following this line of argument and based on lactate dehydrogenase (LDH) activity measurements, Sherwood et al. (2002) suggested that overall activity costs should be higher when fish spend most of their time attacking small prey on an anaerobic basis than when they spend most of their time searching with low speed for large prey, an aerobic process. This argument is consistent with the common observations that somatic growth of fish accelerates after diet shifts to larger prey, e.g. from zooplankton to fish diet. Sherwood et al. (2002) showed that the activity of anaerobic enzymes, and hence costly anaerobic energy generation, scales mostly positively with body size. In addition, they found that glycolytic scaling patterns were variable, notably that when a predatory fish underwent an ontogenetic diet shift, the LDH activity decreased sharply. The interpretation was that the number or duration of burst swimming requirements should decline as prey suddenly becomes much larger, because fewer attacks are necessary to achieve satiation. The final result from this interpretation was that a diet shift from small prey sizes (e.g. zooplankton) to much larger prey (e.g. juvenile fish) leads to an energetic advantage that helps to prevent the more than linear increase of activity costs with increasing body size limiting predators growth performance.
We elaborate on this argument to explain the observed increase of PWR in large predators: when a large predator has already undergone a diet shift from small invertebrates to larger fish prey, there is no further possibility of decreasing activity costs by undergoing another sudden diet shift towards a group of even larger prey types, i.e. towards another trophic level. Apart from the fact that there generally is no higher trophic level than fish in the marine aquatic environment upon which fish can feed, there is no such possibility because the abundance of prey fish in the field decreases rapidly with the slope of the biomass spectrum. This decrease is generally not in the same order of magnitude when the diet is shifted from one trophic position to the next, because of the secondary scaling properties of the biomass spectrum (Kerr and Dickie, 2001), i.e. the overall biomass spectrum is build upon a row of biomass domes for each trophic position (phytoplankton, zooplankton, benthos, fish), each of which has a much higher negative slope than the overall spectrum. This means that there can be even higher numbers of small fish in the environment than there are of the largest invertebrates from the next lower trophic position. When a predator already feeds on the highest trophic position of fish prey, there is no way to compensate for the rapidly decreasing prey abundance by foraging on increasingly larger fish. Hence, in order to prevent positively allometrically scaled activity costs becoming a bottleneck for the energy allocation towards growth or reproduction, the growing predators need to limit their energy expenditure for foraging by another mechanism. From a certain body size onwards, they gradually shift their prey preference towards relatively smaller prey sizes. This precludes the predator having to increase its energy expenditure further by increasing the speed and duration of burst swimming chases necessary to capture increasingly large fish prey.
From these combined arguments, we conclude that smaller predators prefer relatively larger fish prey than bigger predators because of their greater ability to generate the energy for a long high-speed chase on a sustainable basis, and to restore their glycogen pools in reasonable time with moderate energy expenditure. Smaller predators gain higher net energy from relatively large fish prey than bigger predators.
The observation of the gradually increasing function of the preferred predator–prey weight ratio is most pronounced in the large demersal predator, cod, which is able to forage on piscivorous prey (e.g. other adult gadoids), but undergoes an ontogenetic diet shift towards fish prey as the dominant food type at larger sizes (approximately 35 cm, depending on the area (Temming et al., 1998)). From 35 cm on, North Sea cod exhibit a sharp increase in the PWR.
A much steeper increase in PWR was observed for grey gurnard, a species that is morphologically adapted to stalk on the seabed, rather than to chasing prey in the pelagic realm and that undergoes an early ontogenetic diet shift from invertebrates to fish (de Gee and Kikkert, 1993). However, grey gurnards are known to ascend from the bottom at night and to hunt for small pelagic fish, e.g. off Jutland, gurnard are major predators on 0-group cod during June and July (de Gee and Kikkert, 1993). From its body shape are grey gurnard can be characterized to be rather a sit-and-wait predator (like a sculpin) than an active forager. Sit-and-wait predators generally have little myotomal red-muscle tissue (Goolish, 1991b), and a 30 cm grey gurnard has a mean red-muscle percentage of only 6% of total muscle mass (Greer-Walker and Pull, 1975). Sit-and-wait predators generally have a large anaerobic scope and rely on costly anaerobic energy production for burst attacks from small body sizes onwards (Goolish, 1991b). Hence, it can be argued that grey gurnard exhibited the resulting steep increase in the PWR, as a consequence of its physiological constraints. It can provide far less aerobically generated "cheap" energy than, for example, an equally sized whiting, for which the red-muscle percentage of 12% is twice as high (Greer-Walker and Pull, 1975). Our results on grey gurnard suggest that the early ontogenetic diet shift to fish prey and the subsequent early increase in PWR with body size lead to maximal foraging efficiency.
For saithe, the limited data available showed a decrease in PWR in quarters 2 and 3, an increase with body size above 45 cm in the first quarter, and a sudden increase in the fourth quarter for the largest size group. This pattern corresponds to the quarterly percentages of fish prey in the diet of the different predator size classes (Temming et al., 1998), which were high when the PWRs increased and vice versa. The red-muscle percentage of 80 cm saithe is low (3%) compared with equally sized cod (17%, Greer-Walker and Pull, 1975). Therefore, we suggest that saithe follows a different feeding strategy to maximize its foraging efficiency. From the very low percentage of aerobic muscle fibres, burst swimming in saithe seems to be exceptionally costly, guiding saithe to rely on high abundances of invertebrate prey during the highly productive seasons when they may be captured on an aerobic basis, because an 80 cm predator does not need burst swimming to capture a 2 cm euphausiid, the main invertebrate prey (Hislop et al., 1997). The suction generated by the opening mouth may be sufficient to swallow the euphausiids during patrolling or streaming, i.e. aerobic swimming behaviour below two body lengths per second (He and Wardle, 1988), the speed associated with foraging in the field (Wyche, 1984). Saithe apparently exhibit a seasonal diet shift to the trophic level of fish prey during winter, when invertebrate abundance is low but small fish, e.g. pearlsides (
5 cm, Maurolicus muelleri) are still available in high numbers.
Our analysis has demonstrated that smaller predators are most likely to gain higher net energy from relatively large fish prey than larger predators. The body size at which a steep increase in PWR occurs is predator-specific and probably a function of the available prey field, i.e. the slope of the size spectrum (Floeter and Temming, 2003). We conclude that bioenergetic individual-based models of large piscivorous fish (e.g. for cod: Krohn et al., 1997) would be more realistic if a strategy to maximize foraging efficiency is taken into account.
|
Acknowledgements |
|---|
This study was partly financed by the European Community under Framework V, contract QLRT-1999-30183 "LIFECO – Linking hydrographic frontal activity to ecosystem dynamics in the North Sea and Skagerrak: Importance to fish stock recruitment".
|
References |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
-
Bundgaard I. and Sparholt H. (1992) Length based multi-species model for estimation of predation mortalities of herring and sprat in the Baltic. ICES Document, CM/D:16. 20 pp.
Chesson J. (1978) Measuring preference in selective predation. Ecology 59:211–215.[CrossRef][Web of Science]
Coull K.A., Jermyn A.S., Newton A.W., Henderson G.I., Hall W.B. (1989) Length/weight relationships for 88 species of fish encountered in the North East Atlantic. Scottish Fisheries Research Report. 43: 81 pp.
Daan N. (1973) A quantitative analysis of the food intake of North Sea cod, Gadus morhua. Netherlands Journal of Sea Research 6:479–517.
Daan N. (1989) Data Base Report of the Stomach Sampling Project 1981. ICES Cooperative Research Report. 164: 145 pp.
Daan N., Bromley P.J., Hislop J.R.G., Nielsen N.A. (1990) Ecology of North Sea fish. Netherlands Journal of Sea Research 26:2–4343–386.
Floeter J. and Temming A. (2003) Explaining diet composition of North Sea cod (Gadus morhua L.): prey size preference vs. prey availability. Canadian Journal of Fisheries and Aquatic Sciences 60:140–150.
de Gee A. and Kikkert A.H. (1993) Analysis of the grey gurnard (Eutrigla gurnardus) samples collected during the 1991 International Stomach Sampling Project. ICES Document, CM/G:14. 25 pp.
Goolish E.M. (1991) Aerobic and anaerobic scaling in fish. Biological Reviews 66:33–56.
Goolish E.M. (1991) Anaerobic swimming metabolism of fish: sit-and-wait versus active forager. Physiological Zoology 64:2485–501.
Greer-Walker M. and Pull A. (1975) A survey of red and white muscle in marine fish. Journal of Fish Biology 7:295–300.[CrossRef][Web of Science]
He P. and Wardle C.S. (1988) Endurance at intermediate swimming speeds of Atlantic mackerel, Scomber scombrus L, herring, Clupea harengus L, and saithe, Pollachius virens L. Journal of Fish Biology 33:255–266.[CrossRef][Web of Science]
Hislop J., Bromley P.J., Daan N., Gislason H., Heessen H.J.L., Robb A.P., Skagen D., Sparholt H., Temming A. (1997) Database Report of the Stomach Sampling Project 1991. ICES Cooperative Research Report, 219. 422 pp.
ICES. (1992) Manual for the international bottom trawl survey - Revision IV. ICES Document, CM/H:3 (addendum). 41 pp.
ICES. (2002) Report of the Workshop on MSVPA in the North Sea. ICES Document, CM/D:4. 82 pp.
Kerr S.R. and Dickie L.M. (2001) The biomass spectrum. Complexity in Ecological Systems Series(Columbia University Press, New York) 352 pp.
Krohn M., Reidy S., Kerr S.R. (1997) Bioenergetics analysis of the effects of temperature and prey availability on growth and condition of northern cod (Gadus morhua). Canadian Journal of Fisheries and Aquatic Sciences 54:113–121.
Pope J.G. and Knights B. (1982) Comparison of length distributions of combined catches of all demersal fishes in surveys in the North Sea and at Faroe Bank. Canadian Special Publication of Fisheries and Aquatic Sciences 59:116–118.
Robb S. (1991) Manual for the ICES North Sea stomach sampling project in 1991. ICES Document, CM/G:3. 12 pp.
Sherwood G.D., Pazzia I., Moeser A., Hontela A., Rasmussen J.B. (2002) Shifting gears: enzymatic evidence for the energetic advantage of switching diet in wild-living fish. Canadian Journal of Fisheries and Aquatic Sciences 59:229–241.
Somero G.N. and Childress J.J. (1980) A violation of the metabolism-size scaling paradigm: activities of glycolytic enzymes in muscle increase in larger-size fish. Physiological Zoology 53:3322–337.
Temming A., Ehrich S., Hislop J., Gislason H., Skagen D., Daan N. (1998) Feeding ecology of the North Sea fish with special emphasis on the data base of the ‘Stomach Sampling Project 1991’ for use in multispecies assessment. Final Report (EU AIR3-CT94-2410) 550 pp.
Underwood A.J., Chapman M.G., Crowe T.P. (2004) Identifying and understanding ecological preferences for habitat or prey. Journal of Experimental Marine Biology and Ecology 300:161–187.[CrossRef][Web of Science]
Ursin E. (1973) On the prey size preferences of cod and dab. Meddelelser fra Danmarks Fiskeri-og Havundersoegelser 7:85–98.
Ursin E. (1975) Food and food preference of the grey gurnard. ICES Document CM/F:4a. 8 pp.
Ursin E. and Arntz W.E. (1985) On the prey size of cod (Gadus morhua) in the western Baltic. ICES Document CM/J:14. 18 pp.
Vanderploeg H.A. and Scavia D. (1979) Two electivity indices for feeding with special reference to zooplankton grazing. Journal of the Fisheries Research Board of Canada 36:362–365.[Web of Science]
Webb P.W. (1975) Hydrodynamics and energetics of fish propulsion. Bulletin of the Fisheries Research Board of Canada 190: 158 pp.
Wyche C.J. (1984) Observations on the behaviour of a saithe (Pollachius virens (L.) school on a temperate reef. Progress in Underwater Science 9:87–98.
Yang J. (1982) An estimate of the fish biomass in the North Sea. Journal du Conseil 40:161–172.[Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||