Abstract

We quantify statistical relationships between hydrography and abundance of Calanus spp. in the Arctic–Atlantic transition zone around Svalbard (78–82°N) during early autumn. The Atlantic species C. finmarchicus was more abundant in warmer and more saline waters, as expected from its distributional core area. Conversely, the Arctic species C. hyperboreus was more abundant in colder and fresher waters. However, the Arctic species C. glacialis showed opposing relationships with hydrography in shallow compared with deep regions. In shallow waters, the numbers of C. glacialis decreased with temperature and salinity as expected, while somewhat surprisingly the opposite trend was found in deep locations. Sub-surface hydrography between 50–150 m was in most cases a better predictor for Calanus spp. abundance than near-surface conditions, the former generally explaining up to ∼50% of the variability in abundance of each species. Despite finding significant relationships between hydrographic properties and the abundance of each of the three Calanus species, we did not detect significant relationships between the total Calanus biomass and temperature.

INTRODUCTION

Recent ocean warming (Levitus et al., 2000) has extended the geographical range of many temperate marine species, although many cold-adapted species have declined (Barry et al., 1995; Holbrook et al., 1997; Beaugrand et al., 2002). There are indications that the Arctic is warming faster than the global mean trend (Manabe et al., 1992; ACIA, 2004). This is partly due to the elevated inflow of relatively warm and saline Atlantic water masses into the Arctic over the last decades (Morison et al., 2000; Hansen et al., 2004). It is, however, not clear how an increase in temperature and salinity (Morison et al., 2000; Schauer et al., 2004) will affect Arctic marine ecosystems.

Pelagic ecosystems, and zooplankton in particular, are likely to be strongly affected by hydrographic changes and variations in water mass distribution and advection (Ottersen and Stenseth, 2001; Hays et al., 2005). This may have ecological and economic repercussions since zooplankton links primary production to higher trophic levels.

Zooplankton production depends strongly on local primary productivity (Irigoien et al., 2004; Richardson and Schoeman, 2004), but there can also be substantial advective transport of organisms between areas and populations (Aksnes and Blindheim, 1996; Heath et al., 1999; Torgersen and Huse, 2005).

Where Arctic and Atlantic water masses meet, the boreal copepod Calanus finmarchicus, the Arctic shelf-associated C. glacialis and the Arctic C. hyperboreus co-occur (Conover, 1988; Hirche and Mumm, 1992). Together, these copepods dominate the mesozooplankton biomass in the North Atlantic and Arctic Ocean (Conover, 1988; Hirche, 1991; Mumm et al., 1998; Melle et al., 2004). Their production supports nearly all higher trophic levels (Bradstreet and Cross, 1982), and they are important for the biological pump of biogenic carbon into the deep ocean (Pasternak et al., 2002).

Here, we assess the links between physical properties in the water column and the abundance of the three Calanus species in a border area between Arctic and Atlantic water masses, as found in the northern Fram Strait and in waters around Svalbard. The inflow of relatively warm and saline Atlantic waters to the Arctic varies strongly between years (Aagaard and Carmack, 1989; Swift et al., 1997; Grotefendt et al., 1998; Saloranta and Haugan, 2001), and a main driver for ocean climate variability in this area is the variable strength of the North Atlantic Current and its northernmost extension, the West Spitsbergen Current (Aagaard et al., 1987). The presence of water masses with differing physical and biological characteristics makes this region well suited for analysing the links between pelagic community structure and physical properties of the water.

Although the biogeography of the Calanus species is well known (e.g. Grainger, 1961, 1963; Jaschnov, 1970; Conover, 1988), this study attempts to quantify the relationships between hydrography and Calanus abundance by using simple statistical models. The objective is to gain new insights regarding the importance of physical variability in governing this important component of the mesozooplankton community.

METHOD

Hydrographic profiles and Calanus spp. were sampled on eight separate surveys during 2000–2004. The dataset was restricted to a period from 23 August–26 September so as to reduce variability caused by seasonal changes in reproduction and population development. The dataset comprised 86 samples collected at 52 locations west and north of Svalbard (latitude 78°54′−82°30′N, longitude 1°12′−22°22′E; Fig. 1).

Fig. 1.

Map of study area with sampling sites (circles). F, G and H indicate the core ranges of C. finmarchicus, C. glacialis and C. hyperboreus. Colours show average water temperature at 150 m depth [annual mean of 2001 (NOAA Atlas NESDIS 45 World Ocean Database 2001. National Virtual Ocean Data System (NVDOS), http://ferret.pmel.noaa.gov/NVODS/servlets/dataset)]. Arrows indicate the West Spitsbergen Current of inflowing Atlantic water. Depth contours are in 50-m intervals down to the shelf break at 500 m (heavy black line), then 500-m intervals to 4000 m.

Fig. 1.

Map of study area with sampling sites (circles). F, G and H indicate the core ranges of C. finmarchicus, C. glacialis and C. hyperboreus. Colours show average water temperature at 150 m depth [annual mean of 2001 (NOAA Atlas NESDIS 45 World Ocean Database 2001. National Virtual Ocean Data System (NVDOS), http://ferret.pmel.noaa.gov/NVODS/servlets/dataset)]. Arrows indicate the West Spitsbergen Current of inflowing Atlantic water. Depth contours are in 50-m intervals down to the shelf break at 500 m (heavy black line), then 500-m intervals to 4000 m.

Temperature (°C) and salinity profiles were measured in situ by shipboard conductivity, temperature and density (CTD) profilers. Zooplankton was collected by vertical net hauls [either a WP2 net: mesh size 180 µm, mouth opening 0.255 m2 or a Multinet (Hydrobios Kiel): mesh size 180 µm, mouth opening 0.25 m2]. Hauls taken with the WP2 net were not depth-stratified; abundance data from depth-stratified hauls obtained with the Multinet were averaged over the entire sampling depth. Half of the stations were located over deep waters (520–3800 m), but for logistical reasons, only the upper 300–400 m were sampled. Depth-stratified hauls from deep stations showed that peak abundances of all three species occurred in the uppermost 100 m, and the population here consisted of younger stages than deeper in the water column. At shallow locations, practically the entire water column was sampled and depth-resolved hauls showed that the populations were more equally distributed throughout the water column. However, peak abundances were observed deeper in the water column where older stages prevailed. Thus, it seems that during the sampling period (late August–late September), the older stages in the Calanus populations had already started the seasonal descent to over-wintering depths (Conover and Huntley, 1991; Hirche et al., 2006). Depth resolved abundance estimates did not reveal clear patterns, possibly because of the reduced sample size (only Mulinet hauls) or the vertical behaviour of Calanus. Thus we present our analysis based on depth averaged abundance data. This should not affect our results for the shallow locations, but for the deeper stations, it may imply a sampling bias against the older developmental stages which would partly be located in their over-wintering habitats below our sampling strata (Conover, 1988; Kaartvedt, 1996). The consequence of this is that our results from the deep locations represent the active parts of the populations in upper waters, and may not be valid outside the sampling period.

Samples were fixed in a 4% formaldehyde-in-seawater solution, brought to the lab and split into random sub-samples (>150 copepodites of Calanus spp.) from which copepodite stages were counted. We distinguished between copepodite stages of C. finmarchicus and C. glacialis as well as young stages of C. hyperboreus from measurements of the prosome length (Daase and Eiane, 2007). Calanus densities (ind m−3) are based on filtered volumes either calculated from flow meters attached to the Multinet or by multiplying mouth opening area of the WP2 net with vertical hauling distance assuming 100% filtration efficiency. Calanus species composition and abundances at the different locations are described in detail elsewhere (Daase and Eiane, 2007).

As a measure for food availability for predators on Calanus, we used estimates of biomass concentration (mg dry weight m−3). For C. finmarchicus and C. glacialis, biomass was calculated from taxon- and stage-specific dry weights (DWs) (Richter, 1994; Hirche et al., 2001) and from weight–length relationships (Hirche, 1991; Mumm, 1991). Average prosome length for each stage was used to convert weight–length relationships into DW estimates. Weight–length relationships given by Mumm (Mumm, 1991) resulted in ash free DWs (AFDW). Those were converted into DWs assuming a ratio of AFDW:DW = 0.9 (Richter, 1994). Species and stage-specific estimates of DWs obtained from these four references were averaged and used to convert abundance of C. finmarchicus and C. glacialis into biomass (DW). DW for copepodite stages CI and CII of C. hyperboreus were obtained in a similar way but in addition values given by Hirche (Hirche, 1997) were included. DW estimates for C. hyperboreus CIII–CVI were taken only from Hirche (Hirche, 1997), since these were season-specific and higher than those given by the other studies.

Measures of local production were unfortunately not available for the season leading up to the copepod sampling. Calanus abundance showed no relation to fluorescence (indicating algal biomass) at the time of copepod sampling (results not shown).

Statistical analysis

Calanus abundance was log-transformed to stabilize the variance. Then, separate linear regressions were made for each species against each hydrographic variable (temperature and salinity) in interaction with a categorical variable indicating whether bottom depth was less or >500 m. Figure 2 demonstrates the approximate homoskedasticity, normality of residuals and linearity of log (abundance) versus the hydrographic predictors. Zero-values were omitted for convenience in the log-transformation; this was a conservative action because all zero-values occurred where the model results predicted low abundance. In addition, we omitted one station from the analysis of C. glacialis because this clear outlier was situated in the extreme southwest of the dataset, far outside the normal range of this species.

Fig. 2.

Relationships between Calanus spp. abundance and hydrographic variables. There are species- and bottom-depth-specific differences in the co-variation between the abundance of each Calanus species and hydrographic variables. Depth categories were “deep” (>500 m, blue) and “shallow” (≤500 m, red). Abundance and hydrographic covariates were averaged from the surface down to 300 m or bottom depth, whichever was less. Regression lines and 95% confidence bands are given for a model allowing mean and intercept to differ between deep and shelf locations, but assuming equal residual variance. Open symbols denote observations omitted from regression: For C. glacialis, the southwestern most station was omitted because it is outside the normal geographical range of this species. For C. hyperboreus, zero-observations were omitted for convenience in log-transformation.

Fig. 2.

Relationships between Calanus spp. abundance and hydrographic variables. There are species- and bottom-depth-specific differences in the co-variation between the abundance of each Calanus species and hydrographic variables. Depth categories were “deep” (>500 m, blue) and “shallow” (≤500 m, red). Abundance and hydrographic covariates were averaged from the surface down to 300 m or bottom depth, whichever was less. Regression lines and 95% confidence bands are given for a model allowing mean and intercept to differ between deep and shelf locations, but assuming equal residual variance. Open symbols denote observations omitted from regression: For C. glacialis, the southwestern most station was omitted because it is outside the normal geographical range of this species. For C. hyperboreus, zero-observations were omitted for convenience in log-transformation.

Abundance data for stations sampled with multiple hauls were pooled to avoid pseudoreplication. Such pooling sacrifices statistical power because it “throws away” some of the data. We therefore also fitted mixed-effect models with a random intercept for each station, and also for each station within each survey. However, the effect estimates were almost identical to those from the pooled-data analyses and we therefore focus only on the simpler model. In addition, we estimated biomass from the pooled abundance data. Log Calanus biomass correlated with log abundance (r was between 0.61 and 0.93 for all species), and hence biomass correlated with temperature and salinity in much the same way as abundance. We therefore limited our analysis to total Calanus biomass in this paper.

The predictors used in the regression analyses (temperature and salinity) were vertically averaged from the surface down to 300 m or the bottom depth, whichever was less. This approach means that the relative influence of the upper waters on hydrographic averages will vary between some data points in the “shallow” category. The depth at which conditions correlated best with vertically-averaged Calanus abundance was identified by regressing abundance on salinity and/or temperature calculated for depths of 0–200 m in 10 m intervals, at 250 m and 300 m.

We also regressed the abundance of individual developmental stages of each species against temperature and salinity and generally found similar trends across stages within each species. For simplicity, we limit our presentation to the pooled abundances for each species.

RESULTS

The relationship between abundance and hydrography differed between the species and depended on bottom depth (Fig. 2, Table I). Calanus finmarchicus correlated positively with vertically averaged temperature and salinity in both deep and shallow locations, whereas the opposite was observed for C. hyperboreus (Fig. 2). The abundance of C. glacialis declined with salinity and temperature in samples from shallow areas, but showed the opposite pattern in northern and deep waters (Fig. 2, Table I).

Table I:

Regression analyses of log-transformed Calanus abundance (as ind m−3 for each species, as well as total DW as mg m−3 for all species) against the Boolean predictor “shallow” (whether bottom depth<500 m) and two hydrological measures (each species–predictor pair analysed separately; six regressions in total)

 Slope Intercept R2 n 
 Deep Shallow Difference Deep Shallow Difference 
Temperature (°C)         
Abundance (ind m−3        
  Calanus finmarchicus 0.79 (0.11)* 0.32 (0.10)** 0.47 (0.15)* 2.38 (0.28)* 3.95 (0.34)* −1.57 (0.44)* 0.62 52 
  Calanus glacialis 0.43 (0.14)** −0.42 (0.12)** 0.86 (0.19)*** 0.80 (0.35)*** 5.40 (0.42)* −4.60 (0.54)* 0.71 51 
  Calanus hyperboreus −0.79 (0.20)* −0.73 (0.19)* −0.06 (0.28)* 2.10 (0.51)* 3.51 (0.62)* −1.40 (0.81)**** 0.44 51 
Biomass (mg DW m−3        
  Total Calanus 0.11 (0.09) −0.09 (0.08) 0.20 (0.12)**** 1.94 (0.22)* 3.99 (0.27)* −2.05 (0.35)* 0.67 52 
Salinity (S-34.75)         
 Abundance (ind m−3        
  Calanus finmarchicus 4.55 (1.06)* 0.64 (0.46) 3.91 (1.16)** 4.58 (0.20)* 5.14 (0.24)* −0.56 (0.32)**** 0.38 52 
  Calanus glacialis 2.81 (1.10)*** −1.94 (0.42)* 4.75 (1.18)* 2.04 (0.20)* 3.44 (0.22)* −1.40 (0.30)* 0.74 51 
  Calanus hyperboreus −5.12 (1.68)** −2.46 (0.70)** −2.67 (1.82) −0.14 (0.32) 0.42 (0.37) −0.57 (0.50) 0.37 51 
Biomass (mg DW m−3        
  Total Calanus 0.59 (0.63) −0.70 (0.27)*** 1.30 (0.69)**** 2.25 (0.12)* 3.46 (0.15)* −1.21 (0.19)* 0.69 52 
 Slope Intercept R2 n 
 Deep Shallow Difference Deep Shallow Difference 
Temperature (°C)         
Abundance (ind m−3        
  Calanus finmarchicus 0.79 (0.11)* 0.32 (0.10)** 0.47 (0.15)* 2.38 (0.28)* 3.95 (0.34)* −1.57 (0.44)* 0.62 52 
  Calanus glacialis 0.43 (0.14)** −0.42 (0.12)** 0.86 (0.19)*** 0.80 (0.35)*** 5.40 (0.42)* −4.60 (0.54)* 0.71 51 
  Calanus hyperboreus −0.79 (0.20)* −0.73 (0.19)* −0.06 (0.28)* 2.10 (0.51)* 3.51 (0.62)* −1.40 (0.81)**** 0.44 51 
Biomass (mg DW m−3        
  Total Calanus 0.11 (0.09) −0.09 (0.08) 0.20 (0.12)**** 1.94 (0.22)* 3.99 (0.27)* −2.05 (0.35)* 0.67 52 
Salinity (S-34.75)         
 Abundance (ind m−3        
  Calanus finmarchicus 4.55 (1.06)* 0.64 (0.46) 3.91 (1.16)** 4.58 (0.20)* 5.14 (0.24)* −0.56 (0.32)**** 0.38 52 
  Calanus glacialis 2.81 (1.10)*** −1.94 (0.42)* 4.75 (1.18)* 2.04 (0.20)* 3.44 (0.22)* −1.40 (0.30)* 0.74 51 
  Calanus hyperboreus −5.12 (1.68)** −2.46 (0.70)** −2.67 (1.82) −0.14 (0.32) 0.42 (0.37) −0.57 (0.50) 0.37 51 
Biomass (mg DW m−3        
  Total Calanus 0.59 (0.63) −0.70 (0.27)*** 1.30 (0.69)**** 2.25 (0.12)* 3.46 (0.15)* −1.21 (0.19)* 0.69 52 

Data are from 52 stations (latitude >78°54′N, longitude <22°22′E, Julian day 237–269). Repeated samples were pooled within each station before this analysis. Entries in the table body give estimates, standard errors in parentheses and statistical significance level (indicated as *0<P < 0.001; **0.001<P < 0.01; ***0.01<P < 0.05; ****0.05<P < 0.1) for the slope and intercept for “deep” and “shallow” stations, as well as for the difference between them.

Table I gives regression analysis (slope and intercept) for “deep” and “shallow” stations, respectively, and the difference in slope and intercept between bottom-depth categories. As an example of interpreting the coefficient estimates, C. finmarchicus in shallow areas increases in abundance by about [exp(0.32)−1] = 38% per °C increase in average temperature. In deep areas, the increase is ∼120%. In contrast, C. glacialis abundance decreases by ∼34% per °C increase in temperature in shallow areas, whereas in deep areas, it increases by ∼53% per °C increase. The interpretation of regressions of total Calanus biomass against hydrography is analogous and all effects on abundance and total biomass are summarised in Table II.

Table II:

Summary of effects (as %change) on the abundance and total biomass of Calanus spp. caused by a one unit increase in temperature or salinity as predicted from regression analyses in Table I

 Temperature (°c) Salinity 
 Deep Shallow Deep Shallow 
Abundance (ind m−3    
Calanus finmarchicus 120 38 9363 90 
Calanus glacialis 54 −34 1561 −86 
Calanus hyperboreus −55 −52 −99 −91 
Biomass (mg DW m−3    
 Total Calanus 12 −9 80 −50 
 Temperature (°c) Salinity 
 Deep Shallow Deep Shallow 
Abundance (ind m−3    
Calanus finmarchicus 120 38 9363 90 
Calanus glacialis 54 −34 1561 −86 
Calanus hyperboreus −55 −52 −99 −91 
Biomass (mg DW m−3    
 Total Calanus 12 −9 80 −50 

Statistically significant effects (at 95% level) are reported in bold font type.

Model fits of vertically averaged Calanus abundance against depth-resolved temperature and salinity were generally as good, or better, at intermediate depths as compared with near-surface waters (Fig. 3). Calanus finmarchicus correlated poorly with salinity but very well with temperature at depths of 0–50 m in shallow parts of our study. In the open sea, the relationship with salinity was as strong as with temperature and best fits were observed with salinity and temperature between 90–180 m.

Fig. 3.

R2 values of regression analyses of Calanus spp. abundance against temperature and salinity calculated for depths of 0–200 m in 10 m intervals, at 250 m and 300 m. Lines show coefficient of determination (R2) versus the depth from which temperature and/or salinity was taken, in regression models with temperature (red line with crosses), salinity (green) and both (black). Note that in the “shallow” category, locations are excluded from regressions with hydrography for depths below their bottom depth. If the excluded locations are outliers in the regression, this can be one reason why R2 might improve with depth (grey shading shows statistical sample size at each depth; measurements at 250 and 300 m were missing for two deep locations).

Fig. 3.

R2 values of regression analyses of Calanus spp. abundance against temperature and salinity calculated for depths of 0–200 m in 10 m intervals, at 250 m and 300 m. Lines show coefficient of determination (R2) versus the depth from which temperature and/or salinity was taken, in regression models with temperature (red line with crosses), salinity (green) and both (black). Note that in the “shallow” category, locations are excluded from regressions with hydrography for depths below their bottom depth. If the excluded locations are outliers in the regression, this can be one reason why R2 might improve with depth (grey shading shows statistical sample size at each depth; measurements at 250 and 300 m were missing for two deep locations).

At shallow locations C. glacialis, and to a lesser extent C. hyperboreus, were better predicted by salinity than by temperature and the relationships with hydrography were best with measurements taken between 50 and 120 m (50–200 m for C. hyperboreus). At deep locations, C. glacialis was more strongly related to sub-surface (100–170 m) than near-surface conditions, and the relationship with salinity was as strong as with temperature (Fig. 3). For C. hyperboreus, temperature had more predictive power than salinity at deep locations and the strongest relationships were observed with hydrographic conditions near the surface (20–80 m). Combining temperature and salinity in the regression models did not improve the regression fits for any of the species compared with the best one-parameter models.

DISCUSSION

The positive correlation of the Atlantic species C. finmarchicus with vertically averaged temperature and salinity and the negative correlation of the Arctic species C. hyperboreus are as expected considering the core distribution areas of these two species and the anticipated effect of mixing water masses (i.e. van Aken et al., 1991; Ashjian et al., 1995; Melle et al., 2004).

For C. glacialis, however, the positive correlation with temperature and salinity in deep locations is somewhat surprising. We find it unlikely that this is an artefact caused by a sampling bias against deep dwelling animals (see Method) as the expected earlier migration to over-wintering depth due to rapid growth in warmer waters (Fiksen and Carlotti, 1998) should counteract the observed relationship. Although C. glacialis is associated with Arctic shelf seas (Grainger, 1963; Conover, 1988), and the central Arctic Ocean, the deeper regions north of Svalbard are outside its main range (Mumm et al., 1998). Thus occurrence of C. glacialis in this area may be ascribed to individuals from core production areas on the shelf being displaced by modified Atlantic water masses (Manley, 1995). Hence, temperature and salinity in this region may act as indicators of advective transport and therefore result in positive relationships with C. glacialis abundance, explaining why C. glacialis is less common under salinity and temperature conditions where it should normally thrive (Grainger, 1963; Conover, 1988; Manley, 1995; Mumm et al., 1998).

Similar effects of advection on zooplankton abundance occur in the Barents Sea (Skjoldal et al., 1992; Tande et al., 2000; Dalpadado et al., 2003; Edvardsen et al., 2003a). For instance, an episode of strong inflow of Atlantic water during the winter of 1982/1983 displaced C. finmarchicus from the southwestern part of the Barents Sea (Skjoldal et al., 1992). The species was scarce in the inflowing water because it spends the winter deep in the Norwegian Sea, below the depth at the entrance to the Barents Sea. Inflow during spring and summer would have the opposite effect.

In simulation studies, advection has been estimated to account for ∼80% of the C. finmarchicus biomass in the southwestern Barents Sea in summer (Edvardsen et al., 2003b). The importance of advection is supported by our observations of Calanus abundance generally being strongly related to hydrography at intermediate depths (Fig. 3). However, surface water conditions seemed more relevant for C. finmarchicus in shallow locations and C. hyperboreus in deep locations, suggesting that their abundance may be more affected by growth conditions and food availability near the surface. Further investigation would benefit from time-series of abundance, production and advection throughout the productive season.

It seems unlikely that Calanus abundance was limited by temperature or salinity per se, as all three species show similar tolerance to low temperature (Hirche, 1987) and because the modest variability in salinity encountered over our study (range: 33.5–35.5) would have negligible effect on plankton physiology (McLaren et al., 1968; Kinne, 1971).

In contrast to our findings, Tande et al. (Tande et al., 2000) found no clear relationship between inter-annual variability in copepod abundance (including C. finmarchicus) and water temperature during spring and early summer in the northeastern Norwegian Sea and Barents Sea. North Atlantic zooplankton distribution, on the other hand, is linked both to temperature and salinity as well as to large-scale circulation indices (Fromentin and Planque, 1996; Heath et al., 1999; Beaugrand et al., 2002) and advective transport (Heath et al., 1999; Harms et al., 2000; Speirs et al., 2006).

Our results suggest that the biomass available for the next trophic level in the pelagic ecosystem is relatively robust to variability in the species composition of Calanus. Calanus finmarchicus dominates numerically when water temperature exceeds 2.2°C in our shallow locations, whereas in the upper layer of the deep locations, it remains the dominant species irrespective of variability in temperature and salinity (Fig. 2). The variability in temperature recorded in our study did not show significant relationships with total Calanus biomass, and salinity resulted in a significant negative relationship only in shallow locations (Tables I and II). However, as C. finmarchicus is smaller than C. glacialis and C. hyperboreus (Hirche, 1991, 1997; Mumm, 1991), an increase in temperature implies that the available energy would be shifted towards smaller individuals. Since prey size affects energy content (Scott et al., 2000) as well as prey detection in fish (Aksnes and Utne, 1997) and birds (Weslawski et al., 1994), the size shift suggested by our models might affect the energy flow in the food web.

The relationship between Calanus abundance and near-surface hydrography was weak in most cases in our study. This is in contrast to reports from lower latitudes, where correlations between zooplankton abundance and sea surface temperature (SST) are often strong (Rutherford et al., 1999; Beaugrand et al., 2002; Lindley and Reid, 2002). A likely reason for this difference is that high heat loss through the surface in the Arctic may alter SST on much shorter time-scales (hours to days) than the time-scales characteristic of population dynamics of these species (months to years).

We have attempted to quantify ecosystem impacts of hydrographic variability which have previously been treated mainly qualitatively and through simulations (Edvardsen et al., 2003b). Our results confirm that advection affects zooplankton abundance, but the relative importance of local production remains to be quantified. Further progress could be made by a comprehensive study covering large areas throughout the productive season, including deep waters and several taxa. This would also provide valuable ground truthing for realistic ocean models coupling water column physics and biology (Wassmann et al., 2006).

FUNDING

Research Council of Norway (CLIMAR 155889/S30) to E.B.

ACKNOWLEDGEMENTS

We thank the captain and crew on R/V Jan Mayen, G. Storvik and T. Schweder for statistical advice, as well as the cruise participants assisting in data collection. The valuable comments of S. Kaartvedt, J. Søreide and two anonymous reviewers on a previous version of the manuscript are very much appreciated.

REFERENCES

Aagaard
K.
Carmack
E. C.
The role of sea ice and other fresh-water in the Arctic Circulation
J. Geophys. Res. C
 , 
1989
, vol. 
94
 (pg. 
14485
-
14498
)
Aagaard
K.
Foldvik
A.
Hillman
S. R.
The West Spitsbergen Current - disposition and water mass transformation
J. Geophys. Res. C
 , 
1987
, vol. 
92
 (pg. 
3778
-
3784
)
ACIA.
Impacts of Warming in the Arctic: Arctic Climate Impact Assessment
 , 
2004
Cambridge
Cambridge University Press
Aksnes
D. L.
Blindheim
J.
Circulation patterns in the North Atlantic and possible impact on population dynamics of Calanus finmarchicus
Ophelia
 , 
1996
, vol. 
44
 (pg. 
7
-
28
)
Aksnes
D. L.
Utne
A. C. W.
A revised model of visual range in fish
Sarsia
 , 
1997
, vol. 
82
 (pg. 
137
-
147
)
Ashjian
C. J.
Lane
P. V. Z.
Smith
S. L.
The Northeast Water Polynya during summer 1992 - distribution and aspects of secondary production of copepods
J. Geophys. Res. C
 , 
1995
, vol. 
100
 (pg. 
4371
-
4388
)
Barry
J. P.
Baxter
C. H.
Sagarin
R. D.
, et al.  . 
Climate-related, long-term faunal changes in a California rocky intertidal community
Science
 , 
1995
, vol. 
267
 (pg. 
672
-
675
)
Beaugrand
G.
Reid
P. C.
Ibanez
F.
, et al.  . 
Reorganization of North Atlantic marine copepod biodiversity and climate
Science
 , 
2002
, vol. 
296
 (pg. 
1692
-
1694
)
Bradstreet
M. S. W.
Cross
W. E.
Trophic relationships at high Arctic ice edges
Arctic
 , 
1982
, vol. 
35
 (pg. 
1
-
12
)
Conover
R. J.
Comparative life histories in the genera Calanus and Neocalanus in high latitudes of the northern hemisphere
Hydrobiologia
 , 
1988
, vol. 
167
 (pg. 
127
-
142
)
Conover
R. J.
Huntley
M.
Copepods in ice-covered seas-distribution, adaptations to seasonally limited food, metabolism, growth patterns and life cycle strategies in polar seas
J. Mar. Res.
 , 
1991
, vol. 
2
 (pg. 
1
-
41
)
Dalpadado
P.
Ingvaldsen
R.
Hassel
A.
Zooplankton biomass variation in relation to climatic conditions in the Barents Sea
Polar Biol.
 , 
2003
, vol. 
26
 (pg. 
233
-
241
)
Daase
M.
Eiane
K.
Mesozooplankton distribution in northern Svalbard waters in relation to hydrography
Polar Biol.
 , 
2007
, vol. 
30
 (pg. 
969
-
981
)
Edvardsen
A.
Slagstad
D.
Tande
K. S.
, et al.  . 
Assessing zooplankton advection in the Barents Sea using underway measurements and modelling
Fish. Oceanogr.
 , 
2003
, vol. 
12
 
a
(pg. 
61
-
74
)
Edvardsen
A.
Tande
K. S.
Slagstad
D.
The importance of advection on production of Calanus finmarchicus in the Atlantic part of the Barents Sea
Sarsia
 , 
2003
, vol. 
88
 
b
(pg. 
261
-
273
)
Fiksen
Ø.
Carlotti
F.
A model of optimal life history and diel vertical migration in Calanus finmarchicus
Sarsia
 , 
1998
, vol. 
83
 (pg. 
129
-
147
)
Fromentin
J. M.
Planque
B.
Calanus and environment in the eastern North Atlantic 2. Influence of the North Atlantic Oscillation on C. finmarchicus and C. helgolandicus
Mar. Ecol. Prog. Ser.
 , 
1996
, vol. 
134
 (pg. 
111
-
118
)
Grainger
E. H.
The copepods Calanus glacialis Jaschnov and Calanus finmarchicus (Gunnerus) in Canadian Arctic-Subarctic waters
J. Fish. Res. Board Can.
 , 
1961
, vol. 
18
 (pg. 
663
-
678
)
Grainger
E. H.
Dunbar
M. J.
Copepods of the genus Calanus as indicators of Eastern Canadian Waters
Marine Distributions
 , 
1963
The Royal Society of Canada, Special Publications No.5, University of Toronto Press
(pg. 
68
-
94
.
Grotefendt
K.
Logemann
K.
Quadfasel
D.
, et al.  . 
Is the Arctic Ocean warming?
J. Geophys. Res. C
 , 
1998
, vol. 
103
 (pg. 
27679
-
27687
)
Hansen
B.
Osterhus
S.
Quadfasel
D.
, et al.  . 
Already the day after tomorrow?
Science
 , 
2004
, vol. 
305
 (pg. 
953
-
954
)
Harms
J. H.
Heath
M. R.
Bryant
A. D.
, et al.  . 
Modelling the Northeast Atlantic circulation: implications for the spring invasion of shelf regions by Calanus finmarchicus
ICES J. Mar. Sci.
 , 
2000
, vol. 
57
 (pg. 
1694
-
1707
)
Hays
G. C.
Richardson
A. J.
Robinson
C.
Climate change and marine plankton
Trends Ecol. Evol.
 , 
2005
, vol. 
20
 (pg. 
337
-
344
)
Heath
M. R.
Backhaus
J. O.
Richardson
K.
, et al.  . 
Climate fluctuations and the spring invasion of the North Sea by Calanus finmarchicus
Fish. Oceanogr.
 , 
1999
, vol. 
8
 
Suppl. 1
(pg. 
163
-
176
)
Hirche
H. J.
Temperature and plankton 2. Effect on respiration and swimming activity in copepods from the Greenland Sea
Mar. Biol.
 , 
1987
, vol. 
94
 (pg. 
347
-
356
)
Hirche
H. J.
Distribution of dominant calanoid copepod species in the Greenland Sea during late fall
Polar Biol.
 , 
1991
, vol. 
11
 (pg. 
351
-
362
)
Hirche
H. J.
Life cycle of the copepod Calanus hyperboreus in the Greenland Sea
Mar. Biol.
 , 
1997
, vol. 
128
 (pg. 
607
-
618
)
Hirche
H. J.
Mumm
N.
Distribution of dominant copepods in the Nansen Basin, Arctic Ocean, in summer
Deep-Sea Res. Pt A
 , 
1992
, vol. 
39
 (pg. 
S485
-
S505
)
Hirche
H. J.
Brey
T.
Niehoff
B.
A high-frequency time series at ocean Weather ship station M (Norwegian Sea): population dynamics of Calanus finmarchicus
Mar. Ecol. Prog. Ser.
 , 
2001
, vol. 
219
 (pg. 
205
-
219
)
Hirche
H. J.
Mulyakshin
S.
Klages
M.
, et al.  . 
Aggregation of the Arctic copepod Calanus hyperboreus over the ocean floor of the Greenland Sea
Deep-Sea Res. Pt I
 , 
2006
, vol. 
53
 (pg. 
310
-
320
)
Holbrook
S. J.
Schmitt
R. J.
Stephens
J. S.
Changes in an assemblage of temperate reef fishes associated with a climate shift
Ecol. Appl.
 , 
1997
, vol. 
7
 (pg. 
1299
-
1310
)
Irigoien
X.
Huisman
J.
Harris
R. P.
Global biodiversity patterns of marine phytoplankton and zooplankton
Nature
 , 
2004
, vol. 
429
 (pg. 
863
-
867
)
Jaschnov
W. A.
Distribution of Calanus species in the seas of the northern hemisphere
Int. Rev. Ges. Hydrobiol.
 , 
1970
, vol. 
55
 (pg. 
197
-
212
)
Kinne
O.
Kinne
O.
Salinity, animals: invertebrates
Marine Ecology, Vol 1: Environmental factors, Part 2.
 , 
1971
London
Wiley-Interscience
(pg. 
821
-
995
.
Kaartvedt
S.
Habitat preference during overwintering and timing of seasonal vertical migration of Calanus finmarchicus
Ophelia
 , 
1996
, vol. 
44
 (pg. 
145
-
156
)
Levitus
S.
Antonov
J. I.
Boyer
T. P.
, et al.  . 
Warming of the world ocean
Science
 , 
2000
, vol. 
287
 (pg. 
2225
-
2229
)
Lindley
J. A.
Reid
P. C.
Variations in the abundance of Centropages typicus and Calanus helgolandicus in the North Sea: deviations from close relationships with temperature
Mar. Biol.
 , 
2002
, vol. 
141
 (pg. 
153
-
165
)
Manabe
S.
Spelman
M. J.
Stouffer
R. J.
Transient responses of a coupled ocean atmosphere model to gradual changes of atmospheric CO2 2. Seasonal response
J. Clim.
 , 
1992
, vol. 
5
 (pg. 
105
-
126
)
Manley
T. O.
Branching of Atlantic water within the Greenland-Spitsbergen passage - an estimate of recirculation
J. Geophys. Res. C
 , 
1995
, vol. 
100
 (pg. 
20627
-
20634
)
McLaren
I. A.
Walker
D. A.
Corkett
C. J.
Effects of salinity on mortality and development rate of eggs of copepod Pseudocalanus minutus
Can. J. Zoo.
 , 
1968
, vol. 
46
 (pg. 
1267
-
1269
)
Melle
W.
Ellertsen
B.
Skjoldal
H. R.
Skjoldal
H. R.
Zooplankton: the link to higher trophic levels
The Norwegian Sea Ecosystem
 , 
2004
Trondheim
Tapir Academic Press
(pg. 
137
-
202.
)
Morison
J.
Aagaard
K.
Steele
M.
Recent environmental changes in the Arctic: a review
Arctic
 , 
2000
, vol. 
53
 (pg. 
359
-
371
)
Mumm
N.
On the summerly distribution of mesozooplankton in the Nansen Basin, Arctic Ocean
Ber. Polarforsch.
 , 
1991
, vol. 
94
 (pg. 
1
-
146
)
Mumm
N.
Auel
H.
Hanssen
H.
, et al.  . 
Breaking the ice: large-scale distribution of mesozooplankton after a decade of Arctic and transpolar cruises
Polar Biol.
 , 
1998
, vol. 
20
 (pg. 
189
-
197
)
Ottersen
G.
Stenseth
N. C.
Atlantic climate governs oceanographic and ecological variability in the Barents Sea
Limnol. Oceanogr.
 , 
2001
, vol. 
46
 (pg. 
1774
-
1780
)
Pasternak
A.
Riser
C. W.
Arashkevich
E.
, et al.  . 
Calanus spp grazing affects egg production and vertical carbon flux (the marginal ice zone and open Barents Sea)
J. Mar. Syst.
 , 
2002
, vol. 
38
 (pg. 
147
-
164
)
Richter
C.
Regional and seasonal variability in the vertical distribution of mesozooplankton in the Greenland Sea
Ber. Polarforsch.
 , 
1994
, vol. 
154
 (pg. 
1
-
90
)
Richardson
A. J.
Schoeman
D. S.
Climate impact on plankton ecosystems in the Northeast Atlantic
Science
 , 
2004
, vol. 
305
 (pg. 
1609
-
1612
)
Rutherford
S.
D'Hondt
S.
Prell
W.
Environmental controls on the geographic distribution of zooplankton diversity
Nature
 , 
1999
, vol. 
400
 (pg. 
749
-
753
)
Saloranta
T. M.
Haugan
P. M.
Interannual variability in the hydrography of Atlantic water northwest of Svalbard
J. Geophys. Res. C
 , 
2001
, vol. 
106
 (pg. 
13931
-
13943
)
Schauer
U.
Fahrbach
E.
Osterhus
S.
, et al.  . 
Arctic warming through the Fram Strait: oceanic heat transport from 3 years of measurements
J. Geophys. Res.
 , 
2004
, vol. 
109
  
C06026, doi:10.1029/2003JC001823.
Scott
C. L.
Kwasniewski
S.
Falk-Petersen
S.
, et al.  . 
Lipids and life strategies of Calanus finmarchicus, Calanus glacialis and Calanus hyperboreus in late autumn, Kongsfjorden, Svalbard
Polar Biol.
 , 
2000
, vol. 
23
 (pg. 
510
-
516
)
Skjoldal
H. R.
Gjøsæter
H.
Loeng
H.
The Barents Sea ecosystem in the 1980s: ocean climate, plankton and caplin growth
ICES Mar. Sci. Symp.
 , 
1992
, vol. 
195
 (pg. 
278
-
290
)
Speirs
D. C.
Gurney
W. S. C.
Heath
M. R.
, et al.  . 
Ocean-scale modelling of the distribution, abundance, and seasonal dynamics of the copepod Calanus finmarchicus
Mar. Ecol. Prog. Ser.
 , 
2006
, vol. 
313
 (pg. 
173
-
192
)
Swift
J. H.
Jones
E. P.
Aagaard
K.
, et al.  . 
Waters of the Makarov and Canada basins
Deep-Sea Res. Pt II
 , 
1997
, vol. 
44
 (pg. 
1503
-
1529
)
Tande
K. S.
Drobysheva
S.
Nesterova
V.
, et al.  . 
Patterns in the variations of copepod spring and summer abundance in the northeastern Norwegian Sea and the Barents Sea in cold and warm years during the 1980s and 1990s
ICES J. Mar. Sci.
 , 
2000
, vol. 
57
 (pg. 
1581
-
1591
)
Torgersen
T.
Huse
G.
Variability in retention of Calanus finmarchicus in the Nordic Seas
ICES J. Mar. Sci.
 , 
2005
, vol. 
62
 (pg. 
1301
-
1309
)
van Aken
H. M.
Quadfasael
D.
Warpakowski
A.
The Arctic Front in the Greenland Sea during February 1989-Hydrographic and biological observations
J. Geophys. Res. C
 , 
1991
, vol. 
96
 (pg. 
4739
-
4750
)
Wassmann
P.
Slagstad
D.
Riser
C. W.
, et al.  . 
Modelling the ecosystem dynamics of the Barents Sea including the marginal ice zone II. Carbon flux and interannual variability
J. Mar. Syst.
 , 
2006
, vol. 
59
 (pg. 
1
-
24
)
Weslawski
J. M.
Stempniewicz
L.
Galaktionov
K.
Summer diet of seabirds from the Franz-Josef-Land Archipelago, Russian Arctic
Polar Res.
 , 
1994
, vol. 
13
 (pg. 
173
-
181
)

Author notes

Communicating editor: R.P. Karris