Depth distribution of Calanus finmarchicus and C . glacialis in relation to environmental conditions in the Barents Sea KIM H. UNSTAD and KURT S. TANDE Unstad. Kim H. & Tande, Kurt S . 1991: Depth distribution of Calanusfinmarchicus and C. glacialis in relation to environmental conditions in the Barcnts Sea. Pp. 409-420 in Sakshaug. E.. Hopkins. C . C. E. & 0ritsland. N. A . (eds.): Procecdings of the Pro Marc Symposium on Polar Marine Ecology, Trondheim. 12-16 May 1990. Polar Research lO(2). Stage composition and vertical distribution of copepodids of Colanus finmarchicus and C. glacialis are described during spring and summer in Atlantic and Arctic waters, respectively. The two species co- occurred in the region of the Polar Front. both in moderate to high population densities. Ontogenetic migration. meaning that the migration range becomes progressively wider with advancing stage. was found in both species. The present study also revealed that C. finmarchicus had modifications in its ontogenetic vertical distribution. The standing crop of phytoplankton, predominantly Phaeocysfis pouchefii. appeared to influence the degree of stage-specific segregation. Both low and high food concentrations tended to increase the vertical distribution of the instars. On thc other hand, a narrow subsurface stratum of abundant phytoplankton led to an aggregation of copepodids at this dcpth. In the region of the Polar Front. where the two species co-occur. C. glacialis had a deeper distribution than C. finmarchicus, thus creating a bimodal vertical distribution pattern within the uppermost 200 m. ' L J ~ P O L A R ~ N S ~ ' ' Kim H . Unstad and Kurf S . Tande. Department of Aquafic Biology. Norwegian College of Fishery Science. Unioersitj of Tromsb. N-9000 Tromsb, Norway. Introduction There is a tendency in pelagic environments for zooplankton and its food supply to be patchily distributed on any spatial or temporal scale (Dagg 1977). Aggregations of zooplankton have been found on horizontal scales of 1-10 km ( K i ~ r b o e 1988), while zooplankton layers, varying in thick- ness from a few to tens of metres have been observed vertically (Vinogradov 1968; Longhurst 1976). The difference in scales is related to the fact that hydro-dynamic and biological variations in the environment are usually greater when con- sidered vertically than when considered hori- zontally. Vertical migration and distribution of zoo- plankton organisms have been the subject of extensive research efforts since the 19th century (Hardy 1971). In the debate concerning this phenomenon, it has been postulated that the explanations have to be considered at two levels (see Vinogradov 1970; Bohrer 1980; Ringelberg 1980). The phenomenon can, from an evol- utionary perspective, be regarded as an adaption to the planktonic way of life (ultimate and teleo- logical mechanisms). On the other hand there is the direct response of the individual, both to environmental and endogenous factors. These can be regarded as the instant and directly acting agents (proximate or causal mechanisms). Most studies concerning vertical zooplankton migrations have been carried out in temperate regions, where the annual variations in the light and temperature regimes are minor, as compared to polar regions. Consequently, the main topics of these studies have been related to die1 verti- cal migrations. In high-latitude systems, which exhibit pronounced seasonal variations in day- length and temperatures, seasonal and onto- genetic vertical migrations are important features of the vertical migration pattern (Bogorov 1946; Longhurst 1976; Buchanan & Haney 1980; Gron- dahl & Hernroth 1986; Eilertsen et al. 1989). The present study is based on an investigation of the vertical distributions of different devel- opmental stages of the calanoid copepod Calanus finmarchicus carried out in 1986 in fjords of north- ern Norway (Tande 1988). The study from these inshore waters clearly points out that C . fin- marchicus in this region exhibited both die1 and ontogenetic vertical migrations during the spring and early summer. Differences in the vertical 410 K . H . Unstad & K . S . Tande distributions of the copepodids were related to differences in environmental conditions (i.e. tem- perature and phytoplankton abundance) and in population densities of C. finmarchicus between the locations. The study suggested that the ver- tical separation could be an important mechanism in reducing the competitive interactions among the copepodids of C. finmarchicus. Based on samples taken at three time periods in May, the response of the different developmental stages to the environmental conditions was found to be continuously modified throughout the most inten- sive growth and recruitment period of C. fin- marchicus. This complex behaviour pattern of C. fin- marchicus in inshore waters could have been mediated by local environmental factors such as topography and tidal action. Therefore, the potential effect of hydrographic variability of coastal water might have obscured the behav- iourally copepod-generated distribution pattern. Thus, in order to readdress this question specifi- cally related to the vertical behaviour of C. fin- marchicus in an anticipated more homogeneous physical environment, a new study was under- taken in offshore waters in the Barents Sea. The time period selected in May and June is con- sidered to be the period of culmination of the spring bloom in Atlantic waters (Tande 1991 this volume). Thus it is likely that the herbivorous zooplankton grazing balances the carbon pro- duction at least in certain areas (Tande & Slagstad 1991). It is expected that the copepod community displays a vertical distribution pattern that leads to an optimal utilisation of the standing stock of phytoplankton (see Lane 1975; Bohrer 1980; Williams & Conway 1980; Tande 1988). The present study was undertaken in order to describe the vertical distribution of C. finmarchicus i n different vertically structured phytoplankton en- vironments; it aimed at delineating any con- sistent pattern between the vertical distribution of C . finmarchicus and the physical and biological environment. The area selected was a transect from 73"N to 75" 30" in the western part of the Barents Sea. The area was sampled twice in the period from 26 May t o 13 June 1987, cor- responding to the same biological period, and with exactly the same methodological approach as the preceding year in inshore waters in northern Norway. Materials and methods Sampling programme Sampling was accomplished during a survey with R / V ENDRE D Y R B Y , from 26 May to 13 June 1987. A series of 3 stations along a transect between 73'00" 28'00'E and 75'30" 30'00'E was taken at the start of the cruise and was repeated at approximately the same locations (except for Sta- tions 6 and 17) at the end of the cruise 14 days later (see Table 1 and Fig. 1). Station time varied slightly, but most of the collections were carried out between 0830 and 1400 CET (Central Euro- pean Time), except at Station 1 (110G1910 C E T ) . Hydrography and light conditions Conductivity, temperature and density measure- ments (CTD) were performed using a Neil Brown Instruments Smart CTD coupled to a Hewlett- Packard computer. In order to construct the iso- plots in Fig. 2, hydrographic measurements from stations at 73'30" 28"OO'E and 74'30" 30'00'E were used, in addition to the sampling stations shown in Fig. 1. Since surface irradiance is Toble I. Timing. position. weather and dcpth data for the diffcrcnt sampling stations ( S t n . ) . Wind speeds were measured at start of sampling. Cloudincss Depth Wind Stn. Date Timc (CET) Position ( ? t ) (m) ( m / s ) 1 26.05 11W1910 73"OO'N 28"O'E W I W 333 4.6 3 28.05 0845-1J00 74"W'N 28"W'E 100 402 9.8 6 31.05 085&1355 74"30'N 30"W'E 9C-100 365 1 .0 13 09.06 0840- 1405 73"OO'N 28"W'E 100 333 4.6 15 11.06 083G1325 74"W'N 28"OO'E 100 102 4.6 17 13.06 0 8 6 1 3 1 5 75"OO'N 30"OO'E 1 OU 387 9.8 Depth distribution of Calanus 411 80°- 7 0'- 76"- Fig. 1 . Map of the Barents 74"- Sea showing the investigated area and the sampling stations (1-17). Also shown are the 72'- average position of the Polar Front (stippled line) and the surface currents: Arctic currents (broken arrows), Atlantic currents (solid arrows) and coastal currents (stippled arrows). 68'- (1987). 70" - approximately inversely proportional to cloud coverage (Kuz'min 1972), data for the latter were used to detect major differences in the light regimes of the sampling locations. Plankton Zooplankton was sampled by means of a pump system (see Solemdal & Ellertsen 1984) which consists of a 90" bent glass fibre tube with a diameter of 40 cm. A plankton net (180 pm mesh) was mounted on the vertically-oriented outlet while the impeller was positioned on the hori- zontally oriented opening. Samples were taken at discrete depths from the surface down to a depth of 195 m, at 15 m intervals. No sampling from the deeper water masses was carried out. Pumping time was set to 6 minutes and water flow was measured with a TSK (Tsurumi-Seiko-Kosha- kusho Co. Ltd.) flowmeter. The flow rate varied, probably related to the position of the pump relative to the currents, and the values typically varied between 5 and 10 m3 min-I. For the collection of phytoplankton, a 5 liter. 20' 25' 30' 35' 40' 45' 50" 5! Niskin water bottle was mounted externally on the pump. The closing mechanism was remotely controlled by the power source of the pump in such a way that the bottle was closed at the initiation of pumping. Both zoo- and phy- toplankton samples were preserved in 2% for- maldehyde in seawater, buffered with borax. 50% (by volume) 1.5 propane diol, which functions as a bacteriacide that also helps keeping setae and appendages soft, was added to the zooplankton conservation medium. In the laboratory, the Calanus specimens were identified and counted using a Wild M3 stereo microscope at either 16 or 4 0 ~ magnification, depending on copepod size. The different stages of Calanus spp. were assigned to species using the prosome lengths given in Table 2. Samples containing large numbers of copepods were split into ten using a Lea-Wiborg splitter (Wiborg 1951). The number of fractions analysed was adjusted so that 50 o r more were counted for each copepodid stage. This procedure gives an estimated value that deviates by less than 30% from the real sample size (Aksnes 1981). From the 412 K . H . Unstad d K . S. Tande 73 0 74 0 75 0 Fig. 2. Isoplots showing temperature (upper) and salinity (lower) for series 1 (left) and series 2 (right), 73 0 74 0 7 5 0 LATITUDE I N l respectively resulting copepod densities, the total population size could be estimated by simple plane integra- tion. Phytoplankton was identified and counted from duplicate subsamples, according to the method described by Utermohl (1958). During this process a Wild Heerbrugg inverted micro- scope was used at 2 0 0 ~ magnification. Results Hydrography and p h y f o p l a n k f o n There was a general decrease in temperature going northwards in the area of investigation in both of the series (Fig. 2 ) . N o thermal strati- fication was seen, except for the most northern area at 75.5"N i n series 1 (Station 6). Isotherms show that temperatures generally decreased between 0.5 and 1.O"C from surface down to a depth of 195 m. in such a way that the differences in surface values between areas tended to persist down to the deepest depth of sampling. Salinity showed little temporal and spatial variation, rang- i n g from 35.15 to 35.25%. A decrease in salinity was found in two regions at 74"N (Station 3) and 75.5"N (Station 6) in series 1, respectively, with low surface salinities (<35.00%0) due to meltwater from sea ice. The phytoplankton communities of all sampling locations were dominated by the prymnesiophyte Phaoecystis pouchetii. Although other phyto- plankton groups were present, the degree of dom- inance of P . pouchefii was so high that this species was considered as representative of the food avail- Table 2 . Calanur app. Identification kcy for C. finmarchicus ( C F I N ) . C. glocialis (CGLA) and C. hvperboreus (CHYP). Modified from Hassel (unpubl.). Prosome length ( m m ) Stage CFIN CGLA CHYP CI C0.85 0.85-0.90 N . 9 0 CII <1 20 1.2& 1.42 > 1 . 4 2 Clll 1 1 . 6 5 1.65-2.15 >2.15 CIV 1 2 . 3 0 2.3W3.00 >3.w CV 1 3 . 0 0 3.0-3 40 >3.40 Females <3.20 3 2 k 4 . 5 0 >4.50 Depth distribution of Calanus 413 20- i o - STAT!ON 1 30 w 0 4 STATION 13 m Fig. 3. Calanus finmarchicur. Stage composition at all stations in the two series able for the copepods. Therefore, the phyto- plankton conditions will be described here with respect to P. pouchetii only. As shown in Table 2, the mean values for cloud coverage varied between 90 and 100%. Thus no great differences in light conditions between the stations could have been caused by variability in cloudness. Calanus finmarchicus: temperature, phytoplankton, and vertical distribution of copepodids. Series 1 . - In the first series, Station 1 exhibited the highest sea surface temperature (3.6"C) com- pared to the other two stations further north (Fig. 4). The abundance of P. pouchetii varied greatly between Stations 1, 3 and 6. At station 1, the lowest phytoplankton standing stock was found. The number of P. pouchetii was highest at 30 m depth, with 1.27 x 106cells I - ' . Below this depth, cell numbers were low. The copepodite stages in the population of C . finmarchicus at Station 1 were represented in relatively equal proportions (see Fig. 3 for stage composition). However, the population, which was estimated to a total size of 8141 individuals m-', contained a somewhat higher proportion (30.2%) of CIII. The depth of STATION 3 STATION 6 STATION 15 n = 66.987 COPEPODIDS STATION 17 maximum occurrence increased with each stage (Fig. 4): CI and CII had their highest numbers at 45 m, CIII at 60 m, and CIV and CV at 75 m and 165 m, respectively. The adult females were most numerous at 180 m. At Station 3 the surface temperature was approximately 3.4"C, steadily decreasing to 2.2"C at 200m depth. The phytoplankton peaked at 75 m of depth, with maximum in P . pouchetii of 2.37 x lo6 cells I - ' . The copepod abundance and stage composition was clearly different from Station 1: CI made up 73% and thus totally dominated in the population of 52,000 individuals m-2. The vertical distribution pattern was almost similar to that at Station 1, except for the tendency of CV and adult females to occupy a shallower depth (see Fig. 4). At the northernmost station (Station 6) the vertical temperature profile was characterised by cold surface water overlying deeper water masses with a temperature of approximately 1.2"C. The bulk of phytoplankton was found between the surface and a depth of 60 m. where the cell num- bers peaked at 30m with 10.4 X 106cells1-'. Here, C. finmarchicus was found at the lowest abundances recorded during the investigation (5300 individuals m-2). A bimodal stage dis- tribution was seen in which the population con- sisted of primarily CV and adult females (see 414 K . H . Unstad & K . S . Tande % 0 20 40 60 80 100 ---% .*..-A 120 200 160 ~ % 0 20 40 60 80 100 % r--- f,,,, I Cell numbers ( x 1 0 6 I - ' ) Stn 6 0 1 2 3 4 Temperature ("C) F i g . 4. Vertical distributions of Calanurfinmarchicus and Phaorytis pouchcrti cell numben. together with temperature profiles at Stations 1 , 3 and 6 . Fig. 3). in addition t o a lesser proportion of the smallest copepodids (CI and CIII). Vertically, CI and CII had their highest numbers at 30 and 45 m. respectively. thus resembling the population dis- tributions at Stations 1 and 3 (Fig. 4). In contrast to all the other sampling locations. the two most advanced instars occupied shallower depths and had a depth of maximum occurrence a t 30 m . Series 2 . - Sampling of the same area two weeks later showed a temperature in the upper water layers which was in general higher than that found at corresponding locations in series 1. A t Station 13 a surface temperature of 3.9"C decreased slightly with depth (Fig. 5). The phytoplankton was present a t notably abundances only in the uppermost 45 m of depth. T h e highest abundance of copepodids throughout the whole survey was found at this station (62,000 individuals m-?). CIII dominated with 45% of the total population. CII and C I V were each represented by almost 20%, while CI constituted t h e smallest proportion of the population (see Fig. 3). A clear ontogenetic vertical distribution was seen, with the majority of the entire population found below the surface maximum in cell numbers (Fig. 5). Depth distribution of Calanus 415 YO % 80 120 n=69.22 1 Stn 13 160 200 - i I l l 1 1 1 n=66.987 Stn 15 n 20 n=51.255 120 160 200 cv ..-*.- Q CIV w 0 1 2 3 4 Temperature (%) Fig. 5. Vertical distributions of Calanusfinmarchicus and Phaocysris poucherii cell numbers. together with temperature profiles at Stations 13. 15 and 17. A surface temperature of 3.6"C was found at Station 15, with a slightly more pronounced ver- tically decrease in temperature compared to the former location. The phytoplankton was found in very high abundance in the uppermost waters, where Phaeocystis pouchetii peaked in maximum abundance (13.9 x lo6 cells 1-I) at 15 m of depth. The total abundance of Calanusfinmarchicus was found at about the same size as that at Station 13 (Fig. 3). The majority of the population appeared as CII, CHI and CIV, and was located in the uppermost 45 m of depth. At the northernmost station surface tempera- tures were approximately 2"C, and decreased to 1.3"C 'at 170 m (Fig. 5). The greatest standing crop ever recorded during the investigation was found at cell densities of approximately 6 x 10' I - ' in the uppermost 6 0 m of depth. The abun- dance of C . finmarchicus was relatively high also at Station 17. Out of a population of 51,300 individuals m-? (Fig. 3), CI and CII made up 36 and 31%, respectively, and thus constituted the largest proportion of the population. Relatively low abundance of CIV and moderately high num- bers of CV and adult females resulted in a bimodal stage-distribution pattern at Station 17, a situation 416 K. H . Unstad & K . S . Tande STATION 6 STATION 1 7 5 n = 6 0 5 3 n = 1 3 6 9 5 COPEPODIDS Fig 6 Cokanns glucrolrr Srage composition dt Stations 6 and 17 which also was seen at Stations 6 and 3. The various stages were distributed vertically from 40 to 8 0 m of depth, although with a tendency for a larger vertical spread, especially pronounced among CV and adult females (see Fig. 5). Calanus glacialis: temperature, phytoplankton and vertical distribution of copepodids. Calanus glacialis was abundant only at the two northernmost sampling locations, Stations 6 and YO 0- 0 20 40 60 80 100 80 40P=.i 120[ % 0 20 40 60 80 100 i - Clll - - - - CIV 13 (see Figs. 6 and 7). The following information i n the data set is emphasised: At Station 6, C . glacialis was present i n low numbers. estimated at 6053 m - ? . The population was composed mainly of two groups of instars, the majority of which were copepodid Stages I and 11, representing50.9 and 21.7%, respectively. CV. which constituted 18.2%, made up the second group. The two early copepodite Stages (CI and CII) were found between 30 and 9 0 m , while the more advanced stages displayed a wider depth distribution, having been recorded at depths between 30 and 105 m (Fig. 7). Low num- bers of CV and adult females were found at all depths, except for a peak in distribution at 30 m , which was somewhat shallower than the younger stages. Samples taken two weeks later and 30 nautical miles further south (Station 17) contained greater amounts of Calanics glacialis and the total popu- lation size was estimated to 13.695 individ- ualsm-2 by plane integration. Of this, the first three copepodite stages made up more than 75% (26.6 CI, 24.9% CII and 23.7% CIII), while the rest of the population was mainly composed of % Cell numbers (x106 I" ) n=6.053 0 2 4 6 8 10 -7- p-- 10.37 I t 7 i I Stn 6 i - i I 1 I I I i k rj M [ Stn 17 0 1 2 3 4 Temperature ("C) Fig. 7. Vcrtical distributions of Calanur glacialir and Phaocpsrij poucltefii cell numhcrr. togcthcr w i t h tcrnpcraturc profilca at Stations 6 and 17. Depth distribution of Calanus 417 copepodids in Stages IV and V. At Station 17, all the copepodid stages of C. glacialis were recorded at all sampled depths (Fig. 7). The tendency for successively larger stages to have shallower depths of maximum occurrence, as could partly be seen at Station 6 , was clearly evident at Station 17. A tendency towards a bimodal depth distribution was also seen, especially in copepodite stage V. Correlation analysis A correlation analysis was performed on untrans- formed depth values of the numbers of each devel- opmental stage. The resulting values were then used in a MDS (Multi-Dimensional Scaling) analysis. The spread in coordinates representing depth distributions of the different populations (Fig. 8) is seen mainly along the horizontal axis (dimension). The largest distance is between the coordinates representing depth distributions for Calanusglacialis at Station 17 and C. finmarchicus at Stations 1 and 6. On the other hand, the coor- dinates for C. finmarchicus at Stations 13, 15 and 17 may be considered to constitute a group. The coordinates representing the vertical distribution pattern of C. finmarchicus at Station 15 are seg- regated from this group, mainly along the vertical axis (dimension 2). Discussion Calanusfinmarchicus, together with its two sibling species C. glacialis and C . marshallae, constitutes a group that is distributed over the entire Atlantic north of 40” latitude, and throughout the Barents Sea, the Polar Basin and the northern Pacific (Jaschnov 1961,1970; Matthews 1969, Fleminger & Kulsemann 1977; Frost 1974). In the Barents Sea, C. finmarchicus is mainly found north of the Polar Front area (see Fig. l ) , while C. glacialis is an indicator species of water masses of Arctic origin and consequently has its main area of dis- tribution north of the Polar Front (see Jaschnov 1961). The area of the present investigation was located in the Atlantic part of the Barents Sea, with the exception of the most northerly area, and C. finmarchicus was the dominating copepod species. At the two stations near the Polar Front, C. glacialis was present in the same order of magnitude as C. finmarchicus. Length ranges of C. finmarchicus and C. gla- cialis have been shown to overlap, and a plasticity 1 - I 16 8 f15 DIMENSION 1 Fig. 8. MDS plot showing coordinates representing the vertical distributions of Colunurfinmurchrcus (f) and C. glociolrs (g) at the different stations (numbers). in body lengths appears when comparing popu- lations from different regions (Frost 1974). The relatively slight degree of morphological diver- gence further complicates distinguishing between the two species when analysing zooplankton samples. Length-frequency distributions that were constructed during the analysis of the present material (Unstad unpubl. results) showed that the prosome lengths of copepodids of the two species showed some degree of overlap. Over- lapping “tails” of the distributions, however, were considered not to give more than marginal effects in distinguishing the two species when they occurred in equal proportions of abundance. At the same time, the use of prosome lengths would lead to a substantial over-estimation of one species if it was represented by a very small frac- tion as compared to the other species. The annual spring bloom in the Barents Sea is characterised by the presence of large amounts of the prymnesiophyte Phaeocystis pouchetii (Zen- kevich 1963). Although inter-annual variations in species composition appear (E. N~st-Hegseth pers. comm.), there is a trend towards Phaeocystis domination in high latitudes, which is especially pronounced for the period after the culmination of the spring bloom (Eilertsen et al. 1981). In the area of investigation, both population size and vertical distribution of P . pouchetii varied and the highest cell numbers were found between depths 418 K . H . Unstad & K . S . Tande of 15 and 60 m. There was also a tendency for the phytoplankton standing crop to be larger at the more northerly locations. The fact that relatively high cell densities were found down towards depths of 150m indicates that the bloom in the area of investigation was in its culminating phase during this study (H. C. Eilertsen pers. comm.). Even though the dominance of Phaeocystis pouchetii may theoretically have been a conse- quence of selective grazing by zooplankton upon other phytoplankton taxa. P . pouchetii is regarded as the main food source for the herbi- vorous zooplankton during the research period. Although the trophic fate of this species has been discussed for years, P . pouchetii is consumed by Calanus finmarchicus and other members of this genus (Turner 1984; Tande & BAmstedt 1987; Hansen et al. 1990b). Huntley et al. (1987) found that P . pouchetii alone could sustain the nutrient demands of C . hyperboreus in terms of carbon. In the life cycle of P . pouchetii, both small, flagellated solitary cells and colonies comprising non-flagellated cells appear. The colonies are often larger than 200 pm in diameter. According to Hansen et al. (1990b), particles with an equiv- alent spherical diameter (ESD) of 50 pm are in the upper proportion of the size range of food particles suitable for Stage 1-111 copepodids of C . finmnrchicus. During the culmination of the spring bloom, the fraction of disintegrating col- onies increases (Hansen et al. 1990b). Colonies in this state may be easier for the small cope- podites to consume. During the period of investigation, the greatest proportion of the C. finmarchicus population inhabited depths from 15 to 100m. Copepodid Stages I and I1 (CI and CII) were generally found at depths of less than 50 m, while CIII and CIV tended to be located above 120 m depths. CV and adult females generally inhabited the upper 100 m of the water column, but the distributions within this depth interval varied between the stations. At one instance (Station 1). the greatest pro- portions of CV and adult females were found at depths of nearly 200 m. The deeper limit of ver- tical distribution can be seen in relation to coexisting copepods; samples from depths greater than 100 m typically contained large amounts of Metridia spp. and Pseudocalanus spp. (Unstad unpubl.). Low numbers of Calanus spp. in the upper 15m of the water column at all stations except Station 15 may be related to illumination preferences. with light intensities near the surface possibly exceeding the light optimum (see Boden & Kampa 1967) of the animals. No samples were taken from depths greater than 195 m, but as very few copepodites were found in the lower part of the sampled depth range this may indicate that no substantial part of the population was located at greater depths. The vertical distribution patterns indicate an existence of ontogenetic vertical migration in C . finmarchicus at Stations 1,13,15 and 17. At these locations, successively older copepodids tended to be located deeper in the water column than the younger developmental stages. N o such tendency was found at Station 6, but this locality differed strongly from the other stations in displaying a stratification where the upper part of the water column showed temperatures lower than the rest of the sampled depth interval. Additionally, C . finmarchicus co-occurred in relatively low num- bers with C . glacialis (see below). The degree of stage-specific segregation seemed to be related t o the vertical distribution of phytoplankton biomass and to the relative abundance of food available t o the copepods, i.e. phytoplankton biomass versus copepod popu- lation size. At Stations 6, 13 and 15, the verti- cal phytoplankton distributions revealed distinct depth-specific peak levels of algal biomass, and the C . finmarchicus populations tended to aggre- gate in these depth strata. At Station 17, the phytoplankton was vertically dispersed and high cell numbers were found from the surface down to depths of 135 m. At this location the degree of stage-specific segregation was greater than at the previously mentioned stations and the depth stra- tum inhabited by each stage was wider. At Station 1, where phytoplankton cell numbers were low throughout the sampled depth interval, the ver- tical dispersion was pronounced. However, this tendency was not as clear at Station 3, even though cell numbers were low also at this location. A small peak in the number of phytoplankton cells at 75 m, together with a high number of copepodid Stages I and I1 at Station 3, are features that distinguish this station from Station 1. It is difficult to conclude that this might account for the difference in dispersion among instars between the two stations. At this point, the present data set is insufficient and further infor- mation is needed. At Stations 6 and 17, the Arctic species C . glacialis was found in the same order of magnitude as the populations of C. finmarchicus. At both Depth distribution of Calanus 419 The present study reveals that C. finmarchicus showed modifications in its ontogenetic vertical distribution. Both low (Station 1) and high food concentrations (Station 17) appeared to increase the vertical distribution of the instars. On the other hand, a narrow subsurface stratum of abun- dant phytoplankton led to an aggregation of cope- podites at this depth (Stations 6, 13 and 15). The observed shifts in ontogenetic vertical distri- butions in C. finmarchicus in high latitude off- shore and inshore (Tande 1988) waters could thus be explained by changes in food availability. Although the present study does not facilitate an examination of alternative causal relationships, similar future studies should encompass the species complexities at the study site, including physical processes and predation. The monitoring of the same body of water for a time period of several weeks is a prerequisite for future inves- tigations of any causal relationships aimed at improving our understanding of vertical behav- iour in zooplankton. stations, CI dominated, but at Station 17, a large proportion of CII-CIV was also found. As for C. finmarchicus at these stations, the copepodids of C. glacialis displayed a depth distribution with a small degree of overlap between the different stages. The ontogenetic tendencies in the vertical distributions of C. glacialis were inverse to those of C. finmarchicus, so that the youngest stages of C. glacialis were found at depths greater than those where advanced stages were found. This contrasts with the findings of Hansen et al. (1990a), where the older stages of C . glacialis were found deeper in the water column than the younger copepodites. However, Williams & Conway (1980) demonstrated an inverse stage- specific vertical distribution in C. finmarchicus as compared to the congenric species C. helgo- landicus in the North Sea and the Celtic Sea, resembling the differences in vertical distributions between C. finmarchicus and C. glacialis found in the present study. At Station 17, the population of C. glacialis was spread more vertically than its sibling species and could be found throughout the whole interval from 15 to 180 m. In this way the bulk of C. glacialis had the deeper distribution of the two species, a tendency similar to what was found by Herman (1983) in Baffin Bay. Even though there was less vertical spread between the two Calanus species at Station 6, similar dif- ferences between the vertical distributions of the two species could also be seen here. The grouping of the coordinates which rep- resent the vertical distribution patterns of Calanus finmarchicus at Stations 3, 13 and 17 by the MDS analysis illustrates their similarity. The increased distance between these points and the coordinate representing the C. finmarchicus profile at Station 15 are most likely related to the upward shift in the latter profile, as compared with the three previously mentioned. The position of the coor- dinate representing the C. finmarchicus profile at Station 1 near the ones representing both this species and C. glacialis at Station 6 may be due to the relatively low representation of both species at these two stations. The differences in onto- genetic distribution patterns between the two sib- ling species are reflected by the position of the coordinates representing the two species at Sta- tions 6 and 17 some distance apart from each other. The relatively large distance between the points representing C. glacialis at the two stations may be related to the differences in population sizes of this species. Acknowledgements. -The authors would like to thank B . Vaaja for analysing the phytoplankton material. Constructive criticism from anonymous referees is also appreciated. References Aksnes, D . L . 1981: Undersekelse av zooplankton popu- lasjonsdynamikk i Lindhspollene. 1979. Cand. real. thesis. University of Bergen. 122 pp. Boden, B . & Kampa. M . 1967: The influence of natural light on the vertical migrations of an animal community in thc sea. Symp. Zool. SOC. Lond. No. 19, 15-26. Bogorov, B . G. 1946: Peculiarities of diurnal vertical migrations of zooplankton in polar seas. J . M a r . Res. 6, 2 6 3 2 . Bohrer, R . N. 1980: Experimental studies on die1 vcrtical migration. Pp. 111-121 in Kerfoot, W. C. (ed.): Eoolurion and ecology of zooplankton communities. University Press of New England. Buchanan, C . & Hancy. J . 1980: Vertical migrations of zooplankton in the Arctic: A test of the environmental controls. Pp. 6%79 in Kerfoot, W. C. ( e d . ) : Eoolution and ecology of zooplankton communities. University Press of New England. Dagg, M . 1977: Some effccts of patchy food environment on copepods. Limnol. Oceanogr. 22, 99-107. Eilertsen. H. C.. Schei. B . & Taascn. J . P. 1981: lnvestigations of the plankton community of Balsfjorden. northern Norway. The phytoplankton 1976-1978: Abundance, species com- position and succession. Sarsia 66, 126-141. Eilertsen. H. C . . Tande. K. S . & Taasen. J . P. 1989: Vertical distributions of primary production and grazing by Colanus glacialis Jaschnov and C . hyperboreus Kreyer in Arctic waters (Barents Sea). Polar Biol. 9, 253-260. 420 K . H . Unstud & K . S . Tande Fleminger, A. & Hulsemann. K. 1977: Geographical range and taxonomic divergence in North Atlantic Calanus f C . helgolandicus. C. finmarchicus and C. glacialis). Mar. Biol. 40. 233-238. Frost, B. W. 1974: Calanusmarshallae. a new speciesofcalanoid copepod closely allied to the sibling species C. finmarchicus and C . glacialis. Mar. Biol. 26. 77-99. Grondahl. F. & Hernroth, L. 1986: Vertical distribution of eopepods in the eurasian part of the Nansen Basin. Arctic Ocean. Pp. 311-320 i n Schriever. G , Schminke. H . K . & Shih. C . T. (eds.): Proceedings of the 2nd. Inr. Conference on Copepoda, Oftawa. Syllogeus No. 58, Ortawa. Hansen. B., Berggreen. U C.. Tande. K. S. & Eilertsen. H. C. 1990a: Post-bloom grazing by Calanus glacialis. C . finmarchicus and C . hyperboreas in the region of the Polar Front. Barents Sea. Mar. Biol. 104. 5-14. Hansen. B., Tande. K. S. & Berggreen. U. C. 199Ob: On the trophic fate of Phaeocystis pouchetii (Hariot). 111. Functional responses in grazing demonstrated on juvenile stages of Cal- anus finmarchicus (Copepoda) fed diatoms and Phaeocystis. J . Plankron Res. 12. 1173-1187. Hardy. A. C . 1971: The Open Sea. Part 1. The norld of plank- ton. 2nd. ed. Collins. 335 pp. Herman, A . W. 1983: Vertical distribution patterns of copepods. chlorophyll and production in northeastern Baffin Bay. Limnol. Oceanogr. 28. 709-719. Huntley. M., Tande. K . S . & Eilertsen. H . C . 19x7: On the trophic fate of Phaeocysris poucherii (Hariot). I I . Grazing rates of Calanus hyperboreus ( K r ~ y e r ) on diatoms and dif- ferent size categories of Phaeocystis pouchetii. 1. Exp. Mar. Biol. Ecol. 110, 197-212. Jaschnov. W . A . 1961: Water masses and plankton. 1. Species of Calanus finmarchicus s.1. as indicators of definite water masses. Zool. Z u r n a l 4 0 . 1314-1334 (Translated from Russian by N. F. C. Owen. Scottish Mar. Biol Ass. Oceanogr. Lab.. Edinburgh. 1961). Jaschnov. W . A . 1970: Distribution of Calanus species in the seas of the northern hemisphere. Int. Reoue. Ges. Hydrobiol. 55. 197-212. Kiarboe. T. 1988: P l a n k t o n f ~ d e k ~ d e n : Bioenergetiske og oko- logiske studier. ISBN 87-7481-550-4. Danmarks Fiskeri- og Haoundersmgelser. 51 pp. Kuz’min. P. P. 1972: Melting of snow cover. Israel Progr. Scr. Translations. Jerusalem. Lane. P . 1975: The dynamics of aquatic systems: A comparative study of the structure of four zooplankton communities. Ecol. Monogr. 45, 307-336. Longhurst. A . R. 1976: Vertical migration. Pp. 116-139 in Cushing. D. H. & Walsh. J . J . (eds.): The Ecology o f t h e Seas. Blackwell. Oxford Matthews. J . B. L. 1969: Continuous plankton records: The geographic and seasonal distribution of Calanusfinmarchicus S. L. in the North Atlantic. Bull. Mar. Ecol. 6, 251-273. Midttun, L. & Loeng. H. 1987: Climatic variations in the Barents Sea. Pp. 113-127 in Loeng, H . (ed.): The effect of oceanographic conditions on distribution and population dynamics of commercialfish stocks in the Barents Sea. Proc. 3rd Soviet/Nonvegian Symp.. Murmansk 2 6 2 8 May 1986. Inst. Mar. Res.. Bergen. Ringelberg, J . 1980: Introductory remarks: Causal and teleo- logical aspects of diurnal vertical migration. Pp. 6 5 4 8 in Kerfoot, W. C. (ed.): Eoolution and ecology of zooplankton communities. University Press of New England. Solemdal P. & Eliertsen, B. 1084: Sampling fish larvae with large pumps; quantitative and qualitative comparisons with traditional gear. Flodeuigen rapporfser. 1. 335-363. Tande. K. S. 1988: An evaluation of factors affecting the vertical distribution among recruits of Calanus finmarchicus in three adjacent high-latitude localities. Hydrobiologia 167/168. 115-126. Tande. K. S. 1991: Calanus in North Norwegian fjords and in the Barents Sea. Pp. 389-407 in Sakshaug. E.. Hopkins. C. C. E. & Britsland. N. A . (eds.): Proceedings of the Pro Mare Symposium on Polar Marine Ecology. Trondheim. 12-16 May 1990. Polar Research 10(2). Tande. K. S. & Bsmstedt. U . 1987: On the trophie fate of Phaeocysris pouchetii. 1. Copepod feeding rates on solitary cells and colonies of P. pouchetii. Sarsia 72, 313-320. Tande. K . S. & Slagstad. D. 1991: Regional and interannual variations in biomass and productivity of Calanus fin- marchicus in subarctic environments. Oceanologica Acta. I n press. Turner. J . T . 1984: The feeding biology of some zooplankters that are important prey items of larval fish. N O A A Techn. Rep. NMFS7. 28 pp. Utermohl. H . 1958: Zur vervollkommnung der quantitativen Phytoplankton-Methodik! M i t t . Int. Verein. Theor. Angew. Limnol. 9. 1-38. Vinogradov. M. E. 1968: Vertical dhtribution of the oceanic zooplankton. Nauka Publ. House. Moscow (translated from Russian to English. 1970. Israel Progr. Sci. Translations, Jerusalem. 320 pp.). Wiborg. K. F. 1951: The whirling vessel. - An apparatus for fractioning of zooplankton samples. Fiskeridir. S k r . Ser. Haounders. 9, 1-16. Williams. R . & Conway. D . 1980: Vertical distributions of Calanus finmarchicus and C . helgolandicus. Mar. Biol. 60, Zenkevich. L. 1963: Biology of the seas ofthe USSR. Allen and 57-6 1. Unwin. 995 pp.