Estimated food consumption of minke whales Balaenoptera acutorostrata in Northeast Atlantic waters in 1992-1995

Data on energy requirements, diet composition, and stock size were combined to estimate the consumption of various prey species by minke whales (Balaenoptera aeutorostrata) in Northeast Atlantic waters. In the period 1992-1995, the stock of 85,000 minke whales appeared to have consumed more than 1.8 million tonnes of prey per year in coastal waters off northern Norway, in the Barents Sea and around Spitsbergen during an assumed 6 month stay between mid-April and mid-October. Uncertainties in stock estimates suggest a 95% confidence range of 1.4 2.1 million tonnes. The point estimate was composed of 602,000 tonnes of krill Thysanoessa spp. , 633,000 tonnes of herring Clupea harengus, 142,000 tonnes of capelin Mallotus villosus, 256,000 tonnes of cod Gadus morhua, 128,000 tonnes of haddock Melanogrammus aeglefinus and 54,500 tonnes of other fish species, including saithe Pollaehius virens and sand eel Ammodytes sp. Consumption of various prey items by minke whales may represent an important mortality factor for some of the species. For example, the estimated annual consumption of herring corresponds to about 70% of the herring fisheries in the Northeast Atlantic in 1995. Minke whale diets are subject to year-to-year variations due to changes in the resource base in different feeding areas. Thus, the regional distribution of consumption of different prey items is highly dynamic. Folkow, L.P. , Haug, T., Nilssen, K.T. and Nord0Y, E.S. 2000. Estimated food consumption ofminke whales Balaenoptera aeutorostrata in Northeast Atlantic waters in 1992-1995. NAMMCO Sci. Pub!. 2:65-80.


Introduction
Attempts to develop multispecies models for the management of marine resources have led to increased interest in the quantitative analysis of the feeding ecology of top predators.An important top predator in the North Atlantic is the boreo-arctic minke whale Balaenoptera aeutorostrata.The Northeast Atlantic stock, assumed to be one of four minke whale stocks in the North Atlantic, is confined to the waters of Spitsbergen, the Barents Sea, Norwegian coastal waters, the North Sea and other waters off the United Kingdom and Ireland (Anonymous 1977).Part of the Northeast Atlantic stock of minke whales migrates northwards to feeding areas in the Norwegian and Barents Seas 111 spring, and southwards to breeding areas • of unknown location in the autumn (Jonsgard 1966).These animals are reported to feed on various species of zooplankton and fish, particularly herring Clupea harengus, capelin Mallotus villosus and cod Gadus morhua (Jonsgard 1951(Jonsgard , 1982)).In order to obtain quantitative data for evaluation of the ecological significance of predation by minke whales, extensive studies of the energetics and diet of the whales were conducted in a research programme on marine mammals, initiated by Norwegian authorities in 1988.

Ph oto: Pl!r Erik M {jrlellssoll
purpose was to study the energetics and the digestive phys iology of the whales, and in particular, to estimate their energy requirements (see Fo lkow and Blix 1992, Nord0Y et at. 1993, Olsen et at. 1994a, Olsen et al. 1994b, Blix and Folkow 1995, Nord0Y 1995).Although pil ot studies to examine diet were also carried out (Nord0y and Blix 1992), dietary compositi on was not studied in detail until 1992-1994, when the stomach contents of 223 minke whales were analysed, and prey availability was estimated simultaneously (Haug et al. 1995a, Haug et al. 1995b, Haug et al. 1996a, Haug et al. 1997a, Skaug et at. 1997) (Fig. 1).Collection of data for digestive, thermoregulatory, and energetic shldies was also continued (see Martensson et at. 1994, Olsen et at. 1996, Kvadsheim et at. 1996).Fo llowing the termination of scientif ic whaling in 1994, sampling for feeding ecology studies has continued in connection with commercial whaling operations (Haug et al. 1996b, Haug et al. 1997b).
In the present paper, we have combined data on the energy requirements, the diet composition , and the stock size of Northeast Atlantic minke whales (sighting surveys in 1995, see Schweder et at. 1996), to estimate their consumption of va rious prey items.Such info rmation is of importance for assessment of the ecological role of the minke whale, and for f isheries management in Norwegian and adj acent waters.

Modelling strategy
The chemi ca lly bound energy of food eaten by minke whales may be expended (converted to heat or work) through ox idative processes at a rate which is refl ected in the oxygen consumption of the animal, or it can be depos ited as muscles , blubber, visceral fa t and, in pregnant females, a foehls.Additionally, some ingested energy is lost in the urine and faeces.The total daily energy expenditure of free-living minke whales has been estimated from indirect recordings of oxygen consumption rates in freely swimming animals (Blix and Folkow 1995), and consequently include energy costs fo r maintenance, locomotion, thermoregul ation, excess postprandial heat production, maternal costs of gestation etc.The amount of energy deposited through tissue growth has been estimated for males and females of various age groups, from diffe rences in the masses and energy densities of various tissues in whales sampled during spring and autumn (22 and 42 individuals, respectively;  Nord0Y et at.1995).Energy costs of lactation were not included, since the majori ty of minke whal e calves are weaned before arriving in Norwegian and adj acent waters (Jonsgard 195 1).The expended and deposited energy, the metabolizabl e energy (ME), is obtained by intake of an even larger amount of energy, the gross energy intake (GEl), the difference between the two being made up of energy lost in urine (ca.8%, Lavigne et at. 1982) and faeces (ca.8%, Nord0y et at. 1993, Martensson et at. 1994).The GEl equals the energy requirements of the animals.

-
Christensen 1981) and the rest, which were classified as immatures.Based on the size distribution of 223 animals caught during the 1992-1994 scientific whaling programme (see Haug et ai. 1997a), 24% ofthe whales were assumed to be mature and 76% immature.6.According to Christensen (1981), ca.95% of all sexually mature females of the Northeast Atlantic stock of minke whales are pregnant annually.In our calculations, we have assumed that all physically mature females were pregnant.Based on the length distribution of immatures (Haug et ai. 1997a) and unpublished data on body length vs pregnancy rates for 50 females (Erling S. Nord0Y, unpublished data), we calculated that 25% of the physically immature females were also pregnant.7. Based on a length -mass relationship for minke whales caught in July/August (Folkow and Blix 1992), physically mature whales (assigned a body length of 2:8 m, under assumption 5) were predicted to have a body mass which averaged 5,900 kg (95 % CI, 4,800 to 7,000 kg) during the modelling period.The average body length of immatures of 7 m (N ord0Y et af.1995) was used in a similar way to predict an average body mass of 3,8 00 kg (95 % CI, 3,000 to 4,400 kg) for this group.Further, we assumed that the daily energy expenditure of matures was 80 kJ/kg (Blix and Folkow 1995), regardless of sex and season, i.e. , 470,000 kJ/day for adults.The value was determined based on recordings of respiratory rates (Blix and Folkow 1995) and lung volumes (Folkow and Blix 1992) of minke whales, and on published respiratory data (tidal volume as a fraction of lung volume, and the fraction of oxygen in the lung that was taken up by blood) for other cetacean species (e.g. , Wahrenbrock et af.1974).The coefficient of variance (cv) of the value was estimated to be 0.25.Immatures were assumed to maintain elevated basal metabolic rates (BMR) due to growth.Growing mammals in general appear to maintain resting metabolic rates that are approximately twice as high as the BMR predicted according to Kleiber (1975) (e.g., Lavigne et al. 1986).The one existing study of metabolic rates in young, growing baleen whales (grey whales Eshrichtius robustus, see Wahrenbrock et af. 1974) indicates that this may be true also for these mammals.Therefore, we have assumed that immature minke whales maintain a field metabolic rate which corresponds to that of adults (i.e. , 80 kJ/kg), plus a value corresponding to the BMR according to Kleiber (1975) , to account for the fact that they were growing.In doing this, the resulting estimate of average daily energy expenditure of immatures was found to be 445,000 kJ/day (estimated cv=0.25).8.The average increase in muscle mass due to growth / replacement of muscle and deposition offat in muscle, was assumed to be 350 kg (95% CI, 155 to 550 kg) for matures and 247 kg (95% CI, 100 to 400 kg) for immatures over a period of 112 days (based on data by Nord0Y et al. 1995).The resulting growth rate was extrapolated to apply for a period of 180 days .Energy densities of muscle samples were lower in spring (5.4 kJ/g) than in summer and autumn (6 .0 and 7.0 kJ/g, respectively; Nord0Y et af.1995), and we assumed that less energy was deposited as fat, and more as re-growth of ske leta l muscle, in spring than 111 summer/autumn.Given the lower energy density of muscle fibers than of fat, the proportional increase in muscle mass would be expected to be higher in spring than during the rest of the feeding season, and 40% of the increase in muscle mass was therefore assumed to take place in spring, while the corresponding va lues for summer and autumn were both assumed to be 30%.9. Blubber deposition was assumed to amount to 208 kg (95% CI, 140 to 275 kg) in matures and 136 kg (95% CI, 80 to 195 kg) in immatures over a period of 112 days (based on data by Nord0Y et af.1995).The resulting rate of deposition was extrapolated to apply for a period of 180 days, and the deposition of blubber was assumed to take place with 20% occurring during spring, 30% during summer, and 50% during autumn.This was based on the assumption that more energy is deposited as muscle growth, and less as blubber, in spring than in autumn, and also that most prey species were more energy-rich in autumn than in ------------------------------------------------------spring.The energy densities of deposited blubber were set at 27.5 ± 2.5 kl/g during spring, and 30.6 ± 3.0 kl/g during summer and autumn (Nord0Y et al. 1995).10.Visceral fat deposits were insignificant in animals caught in spring, but were substantial in animals caught in autumn.The visceral fat deposition was assumed to amount to 94 kg (95% CI, 75 to 11 2 kg) in matures and 63 kg (95 % CT, 47 to 79 kg) in immatures, over a period of 11 2 days (based on data by Nord0Y et al. 1995).The resulting rate of deposition was extrapolated to apply for a period of 180 days, and deposition of visceral fat was assumed to take place with 40% occurring during sunmler and 60% during autumn.The energy density of deposited visceral fat was assumed to be similar to that of blubber, i.e., 30.6 ± 3.0 kl/g during both summer and autumn (Nord0Y et al. 1995) .II .Foetal growth was assumed to result in a foetal body mass of 45 kg (95% CI, 38 to 53 kg) in mid-October (Nord0y et al. 1995), and the exponenti al growth was assumed to take place with 10% occurring in spring, 30% in summer and 60% in autumn.The energy density of the foetus was assumed to be constant (3.80 ± 0.14 kl/g) during growth (Nord0y et al. 1995).12.The sum of the energy expenditure and the energy deposited in muscle, blubber, visceral fat and foetus corresponds to the metabolizable energy (ME) .ME was assumed to represent 92% of the digestible energy (DE), the remaining 8% being lost in the urine (Lavigne et al. 1982).DE, in turn, represents 92% of the gross energy intake (GEl), if assuming 8% of GEl was lost in the faeces (Nord0Y et al. 1993, Martensson et al. 1994) .13. Data on seasonal changes in energy densities of prey were taken from Martensson et at. (1996).The energy densiti es of sa ithe Pollachius virens and haddo ck Melanogrammus aegleflnus, for which energy density data do not exist, were assumed to be similar to those of cod.14.Finally, we assumed that all seasonally varying parameters (energy density of prey, blubber deposition, muscle growth, visceral fat deposition, foetal growth) changed in the same manner, regardless of latitude.

Determination of relative diet composition
Diet composition, classified according to subarea and season, was estimated from data collected during sampling in 1992-1994 and 1995 (Haug et al. 1995a, Haug et al. 1995b, Haug et al. 1996a, Haug et al. 1996b).The 1992 diet data from area ES were excluded from the analyses due to the collapse in the capelin stock between 1992 and 1993 (see Hamre 1994, Gj0sreter 1995).All diet composition data were based on reconstructed prey biomass in minke whale forestomachs (see Haug et al. 1995a, Haug et al. 1996a), and the prey organisms were grouped into the following 8 taxa: krill (Thysanoessa spp.), herring, capelin, cod, haddock, saithe, sandee l (Ammodytes sp.) and others.The diet composition was presented as percentage mass of each prey group using the individual mass index: where b ij is the estimated biomass of prey group i in whale number), b) is the total mass of all prey groups in whale number) , and n is the total number of examined whales.Recorded masses of individual minke whale forestomach contents vary considerably (0-250 kg), presumably in relation to the feeding and digestive phases in which the whales were caught (see Haug et al. 1997a).By using this individual mass index, each forestomach is given the same importance irrespective of the prey mass contained.

Modelling procedure
The modelling procedure largely followed that outlined by Nord0Y et al. (1995): Estimates of ME were made for four classes of animals (physically mature males, physically mature (pregnant) females, growing pregnant females, imrnatures) and for each of the three seasons (spring, summer and autumn) (Tabl e 2).These ME values were then multiplied by the numbers of animals of each class within each of the three defined subareas.The ME values for all classes were summed for each season and area and the sum s were converted to GEl values as described.

RESULTS
Energy expendi ture (for maintenance, locomotion, thermoregulation etc.) constituted the bulk (76-93%) of the energy requirement of whales in all groups (Tabl e 2).During spring > 90% of ME was expended for these purposes in all groups.
The proportion of energy deposited in foetus by pregnant females was very small (0.05 -0.3%).
The maj or proportions of energy storage occurred as blubber deposition, particularly in autumn when 56-58% of stored energy was deposited as blubber.In general, energy requirements were higher in autumn than in spring and summer.
Estimates of energy requirements (Table 3) and food consumption (Table 6) vari ed considerably between subareas, along with diffe rences in minke whale abundance in di ffe rent subareas.Thus, food consumption was greatest in subarea EB, due to the large number of whales present.
Assuming a point estimate of 84,76 1 minke whales in the subareas ES, EB and EC, their total food consumption in the period between 15 April and 15 October was estimated to amount to more than 1. 8 million tonnes, of which 602,000 tonnes was krill , 633 ,000 tonnes herring, 142,000 tonnes capelin , 255,000 tonnes cod, 128,000 tonnes haddock and 54,500 to nnes other fish species, including saithe and sandeel (Table 6, Fig. 3).
The prey composition varied considerably both between periods and geographical subareas.The consumed biomass was larger in spring than in both summer and autumn.In the northernmost subarea (ES), the diet consisted mostly of krill, particularly during spring and summer when this food item made up 85% to 88% of the biomass consumed by the whales (Table 4).Capelin and cod were also important prey items in ES, at least during the autumn.Di et compos ition appeared to be more variable in subareas EB and EC than in subarea ES.Herring appeared to be particularly important in subareas EB and EC, but while immature fish were taken in subarea EB, the whales in subarea EC consumed mature herring (Tables 3 and 5).In summer and autumn, herring constituted 58% to 96% of the consumed biomass in these two subareas (Table 6).In subarea EB, krill and capelin were taken in large amounts during spring, and sandeel during summer.Cod and haddock were consumed in considerable quantities in all periods.

DISCUSSION
Minke whales of the Northeast Atlantic represent one of the most euryphagous stocks of baleen whales (Haug et al. 1995a, Haug et al. 1996a), but the bulk of their diet is comprised of relatively few species.Krill and herring, the two most prominent prey items in the diet, were consumed in approx imate ly equal amounts.
Together, these two prey species accounted for 68% of the total bi omass eaten.Gadoids (cod, haddock and sa ithe) represented 2 1 %, and capelin 8% of the consumed bi omass.Recent studies of minke whales have revealed dietary heterogeneity between years, presumably as a result of changes in prey resources in the feeding areas of the whales (Haug et al. 1995b, Haug et al. 1996a, Haug et al. 1997b).Thus, the temporal distribution of consumption of different prey items is dynamic, and the results presented here represent an mmual average for the period 1992-1995.It is important to emphasise that the presented calculations, yieldi ng a total annual consumption of approximately 1.8 million tonnes of biomass for the period in question, were based on single point estimates fo r several parameters where variati on certainly occurred, including the estimate of minke whale abundance.By including the quantified 95% confidence limits of the abundance estimates (see Schweder et al. 1997) into the present consumption ca lculations, the estimated annual consumpti on by the stock wo uld fa ll within a ra nge of approximately 1.4 to 2. 1 million tonnes of bi omass.
The finding that herring was the species consumed in largest amounts is supported by the results of previous studies suggesting that herring may be the most preferred prey item fo r Northeast Atlantic minke whales (Haug et al. 1996a, Skaug et al. 1997).Simulations run using
Minke whale consumption of herring was almost exclusively confined to subareas EC and EB.
Herring consumed in area EC were adult fish.Numbers of adult herring have increased in this area during late summer and autumn as a result of the gradual rebuilding of the stock after the collapse around 1970 (R0ttingen 1990, R0ttingen 1992).However, only 3% of the total consumption of herring took place in area EC, while 96% occurred in area EB, where immature fish were eaten.The southern Barents Sea has served as the main nursery area for immature herring (O-group and recruits up to 3-4 years old) since 19 88 (R0ttingen 1990, Hamre 1994, Gj0sceter 1995) o Other ffsh (54) recruitment from 1988 onwards have increased the abundance of adolescent herring in minke whale feeding areas in the southern Barents Sea.
The particularly strong 199 1 and 1992 cohorts (Anonymous 1996a(Anonymous , 1996b) may explain the dominant role of immature herring in the minke whale diet in 1992-1 995 .However, during 1995 most of the herring of the 199 1 and 1992 cohorts migrated westwards out of the Barents Sea, and since the 1993-1995 year classes of herring were rather weak (Anonymous 1996b), a reduction in the reliance on immature herring as prey might be expected.Results of dietary analyses carried out on minke whales sampled in 1996 support this (Haug et al. 1997b) : herring was a less prominent component of the di et of Northeast Atlantic minke whales in 1996 than in the years 1992-1995.
Consumption of krill by minke whales was most pronounced in the northernmost area (ES).The prominent role of krill in the northern area seems to be consistent with the current status of the Barents Sea ecosystem: from 1992 onwards there has been a low abundance of capelin and an increase in zooplankton (Anonymous 1996b).

-
, Energy density va lues for haddock and saithe are unavailable -they were set equal to those of cod.

"
Energy density values for sandeel are only avai lable fro m spring.This value was assumed to apply also during the summer and autumn.
-" Minke whale diets also included a small group of va rious prey items for wh ich energy density data were not ava ilable -fo r simplicity these were set equal to the values for sandee!.
a large extent are controlled by predation by the former (Dalpadado and Skjoldal 1996).Capelin appeared to be important as prey for the minke whales in the northernmost area in 1992, whereas, following the collapse of the Barents Sea capelin stock between 1992 and 1993 (Hamre 1994, Gj0sceter 1995), capelin was replaced by krill (Haug et al. 1996b, Haug et al. 1996a).These dramatic changes led us to exclude data for 1992 from the present analyses of food consumption by minke whales in area ES.There is some evidence that krill is a less preferred prey than herring and capelin (Haug et al. 1996a, Skaug et at. 1997), so it is to be expected that the impo11ance of cape lin will increase (from its present consumption level of 142,000 tonnes) as the capelin stock recovers.In 1995 , the Barents Sea capelin stock was at an all-time low level, with a total estimate of 195 ,000 tonnes (of which 120,000 tonnes were maturing) and with very poor year classes being produced in 1993, 1994and 1995(Anonymous 1996b).The Barents Sea capelin has been protected from fisheries since autumn 1993.
Analyses of the diet of Northeast Atlantic minke whales have revealed that cod and haddock may be less preferred prey than herring and capelin (Haug et al. 1996a, Skaug et al. 1997) .
Nevertheless, large amounts of commercially --------------------------------------------------------73 NAMMCO Scientific Publications, Volume 2 important gadoids are eaten by the whales.Cod seems to serve as an important supplement to the more preferred species, both in subarea EB and ES.The estimated annual consumption of Northeast Arctic cod by minke whales in the period 1992-1995 (255,000 tonnes) was substantjal when compared both with total fisheries (735,100 tonnes in 1995) and estimated total stock biomass (age 3 and older) which was 2 million tonnes in 1995 (Anonymous 1996c).The estimated consumption of haddock was approximately half that of cod (128,000 tonnes), the majority of the haddock being taken in subarea EB.The consumption of haddock was also large compared with the 1995 fisheries and total stock biomass (142 ,500 and 400,000 tonnes, respectively; Anonymous 1996c).The third gadoid species eaten by the whales, saithe, was mainly consumed in subarea EC, but the amounts taken were small.The estimated consumption rates of haddock and saithe are not as accurate as for the other species, since data on energy densities of these species were lacking.We do not think they are far off the mark, however, since these gadoids are very likely to have energy densities close to those of cod, which were the values used in the present study.
Sandeel was consumed in some quantity during the summer, and in the summer of 1992, sandeel was found to be particularly important as food for minke whales in the southeastern parts of the Barents Sea (Haug et ai. 1995a).Sampling could not be carried out in these areas in 1993 and 1994 (Haug et ai. 1996a), possibly leading to some underestimation of the importance of sandeel as prey.
There were large seasonal differences in food consumption by minke whales.Spring was consistently the period of largest biomass intake, due to the low energy density in prey during this period of the year (i.e., the whales must eat larger quantities of biomass to obtain a given amount of energy).By autunm, the transfer of phytoplankton lipids upwards in the food chain (see Falk-Petersen et at. 1990) had contributed to a substantial increase in the energy densities of species at higher trophic levels (Martensson et al. 1996).The apparent lower minke whale feeding rate in summer and fall as compared with spring was, therefore, compensated by the increased prey energy density.In fact, the greatest energy deposition in minke whales occurs late in the feeding season (in fall, see Nress et al. 1998), probably reflecting the time needed for the trophic system to transfer energy from primary producers to top predators.Simi lar patterns, with autumnal deposition of energy (lipids), have been observed for the harp seal Phoca groenlandica, another important top predator in the Barents Sea ecosystem (Nilssen et al. 1997).
We have assumed that the results of the July-August 1995 sighting surveys in the Northeast Atlantic (see Schweder et al. 1997) described the distribution of the whales amongst the three subareas throughout the 180 days feeding period.
Obviously, the distribution of the whales is expected to change from Apri l to October (see Jonsgard 1951, Haug et al. I 996a), but the results of the surveys are the only quantitative information avai lable.The assumpti on of a constant distribution through time obvious ly introduces some bias into the results.For examp le, during the autumna l migration the whales pass through the EC subarea and may feed on the adult herring in the area.As a result, the consumption of adult herring may have been underestimated.
The estimated food consumption rates in the present study are associated with uncertainties other than those mentioned above.For example, our assumpti ons concern ing the duration of the feeding period in northern waters made it necessary to extrapo late from data (e.g. , on energy deposition due to body and foetus growth and fat deposition, as well as on energy densities of prey species) that were co ll ected within more narrow time frames , which introduces some uncertainti es.Moreover, the estimate of energy expenditure used by us was made through indirect calculations of oxygen consumption rates, based on measurements of respiratory rates in freely swimming minke whales, on their lung capacities (determined in newly killed animals), and on literature data on respiratory variables for other cetaceans (B lix and Folkow 1995), all factors obviously being associated with uncertainties.
An assessment of the uncertai nty associated with the energy expenditure estimate is particularly relevant, given its.large influence on the energy requirements of these mammals, and, hence, on the estimated food consumption rates of the stock.The estimated uncertainty was found to be relatively low (CV of about 0.25), which reflects, in particular, the very small variations in respiratory rates observed in these mammals (B lix and Fo lkow 1995).When considering potential sources of errors, it should also be kept in mind that energetic studies of large and unmanageable baleen whales, particularly freely swimming ones, are inherently difficult, for very obvious reasons.We were reassured, however, by the fact that the estimate of energy expenditure rates of adult minke whales was found to correspond to approximately 2.2 times their estimated basal metabolic rate (Blix and Folkow 1995).This va lue is in accordance with the results from much more detailed studies of the energy expenditure of free ly swimming adu lts/subadu lts of other marine mammal (pinniped) species, in which values corresponding to 2 -3 times their basal metabolic rate have been reported (e.g., Lavigne et al. 1982, Markussen et al. 1990, Castellini et al. 1992, Lager et al. 1994).
Moreover, estimated energy requirements for whales belonging to different age and reproductive groups differed by only about 10%, which suggest that our food consumption estimate is not particularly sensitive to potential errors with regard to the grouping of animals.
Erroneous assumptions concerning both diet composition and energy densities in prey species may have biased the results of the calculations, but sensitivity analyses were not performed to assess the possible influence of such errors.Shelton et al. (1997) attempted to quantify uncertainties associated with population size, residency, energy requirements and diet composition in Northwest Atlantic harp seal consumption estimates.They concluded that improved precision in consumption estimates would be obtained by improving knowledge on the diet composition, but this would not necessarily pertain to all prey groups.Uncertainty in population size was the smallest contributor to uncertainty concerning consumption rates (Shelton et al. 1997).To conclude the discussion of uncertain-

--------------------------------------------------------75
ties, we trust that our data represent the best ava il able estimate of the food consumption of this stock, and that, in light of this, our many assumptions are justified.Markussen et al. ( 1992)  Nevertheless, it appears that considerable differences in assumptions may exp lain some of the discrepancy between estimates.Markussen et al. ( 1992) assumed that minke wha les cover 90% of their esti mated annual energy requirements during their summer stay in northern waters (by deposition offat which is then mobilized and utili zed as an energy source during the following winter).A similar strategy has been postulated for other northern hemisphere baleen whales (Lockyer 1987, Vikingsson 1995).Information on daily energy expenditure, the amount of energy deposited as fat, and growth of muscles and foetus seems to indi cate that Northeast Atlantic minke whales would be unable to survive the winter on energy stores built up during the summer alone (see Nord0Y et al. 1995).Our conc lusions imply that these whales also feed on ava ilable sources in their wi ntering areas at lower latitudes and/or that parts of the stock remain on the feed ing grounds at high latitudes for longer periods than the assumed 180 days, for some individuals perhaps even throughout the whole year.Data presented by Vikingsson (1995) show that the extent to which energy is deposited may differ considerably between reproductive classes in fin wha les Balaenoptera physalus, and that these differences may relate to the latitudinal distribution of animals in various seasons.We were unable to detect simi larly large reproductive class differences in energy deposition in Northeast Atlantic minke wha les.Nevertheless, we believe that the food consumpti on rates predicted by Markussen et al. (1992) are overestimated, primarily due to assumptions co ncerning the seasona l migration/feeding strategies of minke whales.
In concl usion, results of the present study suggest that minke whales consume substantial amounts of food in Northeast Atlantic waters, and that their consumption of commercially exploited species such as herring and cod is large enough to be a concern for fis heries management.Refined estimates of the consumption of marine resources by minke whales in the Northeast Atlantic will require the collection of reliable data about the residency and diet of these mammals tlu'oughout the year.
Fig. I: Stolllach cOll tell ts of a lIIillke whale.

-
assumptions be made.

Table 2 .
Estimates of metabolizable energy requirements (ME, the sum of expended and deposited energy, given in kJ) of one individual from each of four groups of Northeast Atlantic minke whales during spring, summer and autumn.See text for explanations of how whale groups and seasons were defined, and of assumptions on wh ich the calculations were based.

Table 3
. Estimates of the energy expenditure (ME) of four different groups of Northeast Atlantic minke whales during spring, summer and autumn in the three subareas ES (Spitsbergen and Bear Island), EB (Barents Sea and coastal aneas of Finnmark and Kola) and EC (Vesternlen and Lofoten).Numbers of whales allocated to each subarea are based on estimates from the 1995 sighting survey.Total gross energy intake (GEl) was calculated from ME data for the whale population in each subarea and season.N = number of whales.All energy values are given in 10" kJ.

Table 4 .
Pooled data on the relative contribution of various prey species (in biomass) to the diet of Northeast Atlantic minke whales in subareas ES (Spitsbergen and Bear Island), EB (Barents Sea and coastal areas of Finnmark and Kola) and EC (Vesternlen and Lofoten) during spring, summer and autumn.The data are based on stomach content analyses of223 whales taken in the period 1992-1995.N = number of whales studied.

Table s .
Energy densities (i n kJ/g) of prey species of ortheas!At lant ic minke whales in spring, summer and autumn.Number of prey samples are given in parentheses.The va lues were derived from data presented by MArtensson e/ af.(1996).

Table 6 .
Estimated prey consumption (in tonnes) of Northeast Atlantic minke whales by prey species, geographical subarea (ES=Spitsbergen and Bear Island; EB=Barents Sea and coastal areas of Finnmark and Kola; EC=VesterAlen and Lofoten) and season.
estimated the food consumpti on of the Northeast Atlantic minke whale stock to be 2.2 million tonlles of biomass.These authors used a simulation model in which they assumed that the consumption took place during a 5 month stay of a stock of 77,000 animals (point estimate) in Northeast Atlantic waters.The lack of quantified uncertainty in both the previous (2.2 million tonlles) and present (1.8 million tonnes) consumption estimate clearly call s for caution in any comparison.