Abstract

We studied the vertical distribution and reproduction of dominant neritic copepod species in the Dogger Bank area and surrounding North Sea to reveal (i) if these species are concentrated in the subsurface chlorophyll maximum layer, (ii) if the chlorophyll maximum offers superior food conditions for reproduction compared with surface waters and (iii) if the secondary production is thus higher in the frontal areas with a subsurface chlorophyll maximum. In addition, we wanted to (iv) identify the most important environmental factors determining the reproduction of neritic copepods in the North Sea. We observed a higher egg production of cultured Acartia tonsa when fed with the seston from chlorophyll maximum, but no evidence of a higher copepod abundance in this layer. Secondary production was highest at the station closest to the upwelling of new nutrients, although seasonal differences in environmental variables probably overrode the differences between frontal and stratified stations. Copepod egg production on an annual basis seemed to be best predicted by the body size and specific fatty acids, with a high egg production, but low hatching success associated with a high EPA:DHA ratio. Total secondary production of small copepods seemed mainly related to the species composition, suggesting that factors controlling abundance of specific species rather than reproduction might be more important in determining the secondary production of copepods.

INTRODUCTION

Coastal areas and many productive frontal areas in the North Sea are dominated by small neritic copepod species, such as Temora longicornis, Pseudocalanus spp., Paracalanus parvus, Oithona spp., Acartia spp. and Centropages spp. (Fransz et al., 1998; Nielsen and Munk, 1998). At the Dogger Bank in the central North Sea, these species can form the majority of zooplankton biomass especially in summer (Fransz et al., 1991; Nielsen et al., 1993), contributing substantially to the diet of larval fish (Nielsen and Munk, 1998). Small copepods are also major grazers of phytoplankton and microzooplankton in the area, with the estimated grazing rates of 15–30% of the daily primary production and a potential to control the biomass of microzooplankton (Nielsen et al., 1993).

The frontal areas of the North Sea, including the Dogger Bank, are characterized by strong subsurface phytoplankton maxima at depths between 15 and 30 m during the stratified season (Richardson et al., 1998, 2000). These subsurface peaks are suggested to be centres for high zooplankton production, where new production of large phytoplankton (mainly dinoflagellates) is efficiently transferred to copepod production (Richardson et al., 1998). The primary production of subsurface peaks over the year is estimated to exceed that of the spring bloom (Richardson et al., 2000), emphasizing the potential importance for small neritic copepod species, with their annual biomass peak in late summer (Nielsen and Munk, 1998). The perceived importance of subsurface chlorophyll maximum for copepod secondary production is based on observations of peak biomass of Oithona spp., Temora longicornis and Pseudocalanus spp. at the pycnocline (Nielsen et al., 1993), as well as a typically higher secondary production at the depth close to the subsurface chlorophyll maxima (Richardson et al., 1998).

Secondary production of copepods is formed by biomass and growth, and mostly measured in terms of egg production. Egg production is influenced by factors such as temperature (Halsband and Hirche, 2001), food concentration (Bunker and Hirst, 2004), composition, mineral quality (Anderson and Pond, 2000) or biochemical quality (Jónasdóttir et al., 1995). Copepod biomass is also influenced by mortality, mainly assumed to be due to predation (Ohman and Wood, 1996). Normal diel vertical migration behaviour of copepods, with an ascent to surface waters for feeding during night time and descent during day time, has been linked to predation risk from visual predators, such as planktivorous fish (Hays et al., 1994, 1996). As the subsurface maximum layer occurs in the deepest part of the photic zone where light levels are low (Richardson et al., 1998), one may assume that vertical migration to greater depths during daytime to avoid visual predators should not be very rewarding. A high copepod biomass in the subsurface chlorophyll maximum layer could thus be due to low predation pressure, while the high production rate could be related to increased food concentration, favourable food composition, superior food quality or low energetic costs of (no) vertical migration.

Our study concentrated on the stratified summer season in the area of the Dogger Bank, although to broaden the range of environmental conditions, we also included transects at other frontal areas in the coastal and central North Sea during different seasons. Dogger Bank is an important feeding ground for planktivorous fish (Daan et al., 1990), and has been studied intensively during the past few decades (Nielsen et al., 1993; Fransz et al., 1998; Richardson et al., 1998), including studies focusing on subsurface phytoplankton blooms (Richardson et al., 2000). We studied the vertical distribution and reproduction of dominant neritic copepod species in this area with three questions in mind. First, we wanted to know if these species are concentrated in the chlorophyll maximum layer, and if they perform diel vertical migration. Second, we wanted to know if egg production rates and hatching success are significantly higher in the chlorophyll maximum layer than in the surface. Third, we wanted to identify the most important environmental factors determining the vertical distribution, egg production and hatching success of neritic copepods in the North Sea. To be able to separate effects of temperature, body size and previous feeding history from the effects of food quantity and quality in the subsurface chlorophyll maximum, we included a bioassay approach (see Müller-Navarra and Lampert, 1996), where cultured copepods were fed with in situ water to estimate their reproduction.

METHOD

Sampling stations

The study took place over four years: 2001, 2002, 2003 and 2005. The experiments in 2001 were designed to study the seasonal succession of in situ egg production and hatching of small copepods (represented by Acartia clausi and Centropages typicus) and their dependence on environmental variables (hereafter referred to as “in situ” experiments), while the sampling in 2002, 2003 and 2005 focused on the late summer vertical distribution and biomass of all dominant neritic copepods. The bioassay study in 2003 was conducted to evaluate the value of subsurface chlorophyll maximum versus surface water for copepod (represented by Acartia tonsa) reproduction (hereafter referred to as the “bioassay” experiment). The sampling stations in 2001 were situated along four transects both in the coastal and central North Sea: T1 was situated in the Skagerrak, T4 in the Dogger Bank area, T5 in the shallow coastal area of the North Sea and T3 at the transition between the Skagerrak and North Sea with both shallow and deep stations. The sampling in 2002, 2003 and 2005 was conducted in the Dogger Bank area (Fig. 1). In all years, the water column was sampled for temperature, salinity and fluorescence profiles (CTD) and microplankton species composition and biomass. In addition, in 2001 and 2003, fatty acid composition and concentration was measured. An overview of experiments, measurements and stations are presented in Table I, while the details of the different methods are described below.

Table I:

Overview of the number of sampling stations, copepod species and measured variables

Measurement Species No. of stations Background 
In situ reproduction A. clausi 22 (2001) T, Chl a, microplankton, FA, depth-integrated copepod biomass (day) 
C. typicus 10 (2001) 
C. typicus 1 (2003) 
Bioassay reproduction A. tonsa 4 (2003); chl max and surface Chl a, microplankton, FA 
Vertical distribution and biomass Dominant neritic species 2 (2002): 5 depths, 9 times a day T, Chl a, microplankton 
3 (2003): 11 depths 
4 (2005): 11 depths 
Measurement Species No. of stations Background 
In situ reproduction A. clausi 22 (2001) T, Chl a, microplankton, FA, depth-integrated copepod biomass (day) 
C. typicus 10 (2001) 
C. typicus 1 (2003) 
Bioassay reproduction A. tonsa 4 (2003); chl max and surface Chl a, microplankton, FA 
Vertical distribution and biomass Dominant neritic species 2 (2002): 5 depths, 9 times a day T, Chl a, microplankton 
3 (2003): 11 depths 
4 (2005): 11 depths 

For the location of stations, see Fig. 1. (T) Temperature, (FA) fatty acids.

Fig. 1.

Map of the study area showing transects and sampling stations in (A) 2001 and (B) 2002, 2003 and 2005. The stations in 2001 were situated along four transects in the coastal and central North Sea and sampled in March–September, while stations in 2002–2005 were all situated at the Dogger Bank area and sampled in July–August. (Cross) stations with Acartia spp. reproduction and abundance (2001; white crosses for March), (circle) stations with Centropages typicus reproduction and abundance (2001; open squares for March), (star) Dogger Bank stations with A. tonsa bioassay (2003) and (open circle) sampling for vertical distribution and biomass (2002, 2003 and 2005). In 2001, the first number refers to the transect and the second number to the station (e.g. 3.3 is the station 3 on the transect 3).

Fig. 1.

Map of the study area showing transects and sampling stations in (A) 2001 and (B) 2002, 2003 and 2005. The stations in 2001 were situated along four transects in the coastal and central North Sea and sampled in March–September, while stations in 2002–2005 were all situated at the Dogger Bank area and sampled in July–August. (Cross) stations with Acartia spp. reproduction and abundance (2001; white crosses for March), (circle) stations with Centropages typicus reproduction and abundance (2001; open squares for March), (star) Dogger Bank stations with A. tonsa bioassay (2003) and (open circle) sampling for vertical distribution and biomass (2002, 2003 and 2005). In 2001, the first number refers to the transect and the second number to the station (e.g. 3.3 is the station 3 on the transect 3).

Hydrography, microplankton and fatty acid analysis

Vertical distributions of temperature, salinity and fluorescence were measured using a SeaBird SBE11 CTD equipped with an in situ fluorometer. Fluorescence profiles were not taken on the Alkor cruise in September 2001 but standard chlorophyll measurements were made at the surface and within the well-mixed water column. Fluorescence was calibrated against chl a, measured in water samples from selected depths and stations during each cruise. The linear regression between fluorescence and chl a typically had an R2 value >0.7 (data not shown).

In 2001, microplankton concentration and composition were measured from the water samples collected at the surface (ca. 4 m), in the chlorophyll maximum layer, and close to the bottom, and integrated over the whole water column. In 2003, phytoplankton was sampled from the surface and from the chlorophyll maximum layer. The water was sampled using Niskin bottles on the CTD rosette. Microplankton samples were preserved in 4% acidic Lugol's solution for phytoplankton and ciliates, and gluteraldehyde for separation of autotrophic versus heterotrophic organisms. Lugol's samples were allowed to settle in a 50-mL settling chamber (Uttermöhl, 1958; Hasle, 1979) and gluteraldehyde samples handled according to Haas (Haas, 1982). Phytoplankton and ciliates were enumerated and geometrical axes of 10 cells per taxon were measured under inverted microscope (Lugol's samples) or epifluorescence microscope (gluteraldehyde samples). Volumes were calculated using the software program Planktonsys 3.11 from BioConsult A/S, and carbon content calculated according to Edler (Edler, 1979) and Mullin et al. (Mullin et al., 1966). A minimum of 400 cells (in total) were counted from each sample: a minimum of 50 cells sample−1 were counted for the most common taxa, while all cells in the 50-mL sample were counted for the less abundant taxa.

Fatty acid concentration and composition in 2001 were measured only in the chlorophyll maximum layer, while in 2003 fatty acids were analysed both from the surface and the chlorophyll maximum layer. In both years, 1.5–5 L of water (depending on seston concentration) were filtered under a low pressure on pre-combusted GF/C (Whatmann) filters, folded into cryovials and immediately frozen at −80°C in nitrogen atmosphere. In 2001, known water volumes (0.5–60 L) were first size fractionated by successive filtration through 80, 40 and 20 µm mesh nylon gauze filters to obtain fatty acid composition in different size fractions. Fatty acids were extracted from filters containing the field collected seston using chloroform:methanol solution (2:1 by volume). A known amount of the fatty acid C23:0 was added to the sample and used as an internal standard for quantification of specific fatty acids. After at least 24 h extraction at −20°C, the chloroform:methanol phase was collected and the filters were washed three times with chloroform:methanol solution. The extracts were washed according to the modified Folch method (Hamilton et al., 1993). After saponification, the samples were transmethylated to fatty acid methyl esters (FAME) using boron trifluoride and then stored in air-tight vials in an argon atmosphere at −80°C until GC analysis. The FAME sample was injected into a gas chromatograph (Hewlett Packard 5809A, with a 30 m omegawax 320 µm column, and equipped with a split/splitless injection system) using helium as a carrier gas at 1.8 mL min−1. FAME were identified based on comparison with retention times of several standards; Larodan polyunsaturated fatty acid (PUFA) standard, fatty acid from the dinoflagellate Prorocentrum minimum to locate 18:5 (n-3), Matreya PUFA-3 and Supelco 18 919, resulting in identification of 42 fatty acids.

Copepod biomass and vertical distribution

The sampling in 2002, 2003 and 2005 was conducted at two to four stations along one Dogger Bank transect, with samples taken at depth-intervals of 10 m (2002) or 5 m (2003 and 2005). The number of daily samplings was one in 2003 and 2005, and nine in 2002. Zooplankton abundance was estimated from samples collected using a zooplankton pump (ca. 1200 L min−1), equipped with a 30 µm net and preserved with Borax buffered formalin (ca. 4% final concentration).

The zooplankton samples were counted from 1/ 4 to 1/64 fraction of the sample, so that the counts exceeded 200 individuals for all dominant copepod species (see Results). Copepods were identified to the genus or species level, and divided into seven developmental stages (nauplii + copepodite I–VI). Loose copepod eggs (not in 2005) and egg-sacs were included in the counts. During 2002 and 2003, all individuals were separated into size classes (with a 25 µm precision) according to prosome (copepodites and adults) or total (nauplii) length, while in 2005 ca. 10 individuals of each species and life-stage were measured. The biomass was calculated from the length–weight regressions of Klein Breteler and Gonzalez (Klein Breteler and Gonzalez, 1982), Sabatini and Kiørboe (Sabatini and Kiørboe, 1994) and Cohen and Lough (Cohen and Lough, 1981) for, respectively, Temora longicornis, Centropages spp. and Pseudocalanus sp., Oithona spp. and Paracalanus parvus. For Microcalanus pusillus, the length–weight regression of T. longicornis was used. Dry weight was converted to carbon assuming 40% carbon content of the dry weight (Mullin, 1969).

To avoid bias due to low number of individuals, the vertical distribution of different life stages was examined in detail only for those species which contributed ≥10% to the total biomass for the given year. To ensure sufficient numbers of individuals, all copepodite stages (I–V) were grouped together. Vertical distribution was expressed both in absolute numbers (ind. m−3) and as a proportion of the population at a given depth (% of total biomass).

Reproduction

In situ egg production and hatching success of Acartia clausi and Centropages typicus were measured along four transects during spring (March–April), summer (May/June, July/August) and autumn (September) 2001. Egg production and hatching experiments were carried out wherever A. clausi and/or C. typicus were present, which resulted in a total of 22 and 10 in situ egg production stations for A. clausi and C. typicus, respectively (Fig. 1, Table I). In addition, in situ egg production of C. typicus was measured in one occasion in 2003. The bioassay in 2003 was conducted at four stations along one transect, which was repeated six times, so that each station was sampled during six subsequent days at approximately same time of the day.

In situ reproduction

Individuals for the in situ experiments were collected by vertical hauls from below the thermocline, with a 180 µm plankton net (WP2) equipped with a 5 L non-filtering cod end. Healthy active females were placed individually into 250 mL bottles containing 64 µm sieved ambient seawater from 4 m depth (15–30 replicates) and the bottles were placed in a dark temperature-controlled room (6–12°C). The sieving removed any ambient copepod eggs, while parts of the small size plankton community passed through. After 24 h, eggs were collected by filtering the samples gently onto a 20 µm mesh and immediately counted using a stereo microscope. The prosome lengths of the females were measured. Eggs from all replicates from a specific station were pooled and incubated in 650 mL bottles containing 20 µm sieved ambient water, at in situ temperature (6–12°C) appropriate for the station where they were collected. After an additional 72 h, the bottle contents were filtered onto a 20 µm mesh, and fixed in 5% Lugol's solution. Eggs, nauplii and empty egg shells were later counted under a binocular microscope. As the incubation time in April–May was too short to obtain complete hatching at the low temperature (6–7°C), hatching success during these months could not be estimated. Weight-specific egg production was calculated based on the length-carbon regression of Berggreen et al. (Berggreen et al., 1988), assuming the carbon content of eggs to be ca. 0.04 µg C egg−1 (Kiørboe and Sabatini, 1995).

Bioassay

The bioassay was conducted using Acartia tonsa, which have been cultured in the National Institute for Aquatic Resources (DTU Aqua) for over 100 generations. Typically, copepods are reared at 18°C and 33‰, and fed Rhodomonas sp. in excess. In the experiments, recently (≤3 days before) matured A. tonsa were fed with in situ water from the chlorophyll maximum layer and surface, as well as with a saturating concentration (>400 µg C L−1) of Rhodomonas sp. (control for standard egg production), at a constant temperature of ca. 15°C. This bioassay approach excluded the effects of body size, maternal nutrition and temperature, so that all the changes in egg production and hatching success were supposedly due to changes in food conditions, either quantity or quality (Müller-Navarra and Lampert, 1996; Koski et al., 2010). However, the use of cultured copepods introduces other problems; such as the typically low egg production, as well as the question of how well do the cultured individuals represent in situ animals (Koski et al., 2010). To reduce these problems, we included the standard food source Rhodomonas sp. to give an estimate of the close-to-maximum egg production of the culture.

The bioassay was run for a total of 6 days, of which the first 24-h period served as an adaptation time. In the beginning of the experiment, ca. 15 adult females and 3 adult males were sorted out from the culture and placed together into each of five replicate bottles of ca. 0.6 L. After each 24-h period, the eggs were counted, the copepod condition checked (dead/alive) and the copepods were thereafter transferred to a new food suspension, which was collected daily. The mortality during the incubations was relatively low, typically <10% day−1 in both surface and chlorophyll maximum water. For the estimation of hatching success, the eggs from the five replicate bottles were pooled into one 0.3 L hatching bottle, which was left to incubate for ca. 3 days. All bottles were placed on a plankton wheel (ca. 1 rpm), in the dark and at ca. 16°C. This was approximately similar to the surface temperature at the study site. At the termination of the hatching experiment, the bottle contents were carefully filtered onto a 50 µm sieve and preserved in 4% acid Lugol's solution, for later counts of eggs and nauplii.

Secondary production

Copepod secondary production was calculated from the depth-integrated carbon biomass and weight-specific egg production, assuming that all stages have equal weight-specific growth rates (Berggreen et al., 1988). This method has many limitations (see e.g. Hirst and Sheader, 1997), and the secondary production rates presented here should thus be taken as rough estimates. To estimate the egg production, we used the egg sacs collected with the zooplankton pump hauls for sac-spawning species, and both eggs collected with the pump and egg production measured in incubations for broadcast-spawning species. The egg production of the sac-spawners was calculated by dividing the depth-integrated egg sac biomass by the biomass of sac-spawning females (egg-ratio method; e.g. Sabatini and Kiørboe, 1994) and temperature-dependent development time of eggs (Andersen and Nielsen, 1997). The egg production of broadcast spawners was first estimated similarly, but since the rates obtained were very low, it seemed likely that the number of eggs in the water column did not accurately represent the egg production (due to e.g. sinking of eggs). We therefore decided to use the regression between egg production and body size (prosome length) obtained from all in situ incubations pooled together to estimate the egg production of broadcast-spawning copepods. However, as the egg production of copepods <1 mm in prosome length was not related to body size (Fig. 2), we decided to use the average egg production of all individuals <1 mm for this size class, and only use the significant regression between prosome length and egg production for the individuals >1 mm (R2 = 0.84; P < 0.0001; Fig. 2).

Fig. 2.

Egg production (eggs f−1 day−1) as a function of prosome length (mm) in all incubations. (Solid circles) Acartia clausi, (open circles) Centropages typicus. The dashed line indicates the significant linear regression between the egg production and body size for >1 mm copepods (see Methods). As the egg production of C. typicus in March and September was exceptionally low (five open circles in the lower right-hand corner of the graph), and clearly out of range in relation to body size, it was excluded from the regression. The regression statistics and the average egg production of <1 mm copepods are indicated in the figure.

Fig. 2.

Egg production (eggs f−1 day−1) as a function of prosome length (mm) in all incubations. (Solid circles) Acartia clausi, (open circles) Centropages typicus. The dashed line indicates the significant linear regression between the egg production and body size for >1 mm copepods (see Methods). As the egg production of C. typicus in March and September was exceptionally low (five open circles in the lower right-hand corner of the graph), and clearly out of range in relation to body size, it was excluded from the regression. The regression statistics and the average egg production of <1 mm copepods are indicated in the figure.

Statistics

Egg production, spawning percentage, brood size, hatching and prosome length were tested for differences between stations or water layers (only bioassay) using one- or two-way analyses of variance (ANOVA); and, after pooling the stations, for differences between months (only in situ reproduction). The Tukey HSD post hoc test was used for pairwise comparisons. If the conditions for normality and equal variance were not met, the non-parametric Kruskal–Wallis test and Dunn's pairwise comparison were used. Pearson correlation analysis was used to check for the correlations between egg production/hatching, body size and temperature, as well as between biomass/secondary production and temperature, chl a and microplankton concentration.

To investigate the dependence of egg production and hatching success on environmental variables, we ran a series of multiple regression analyses. To account for the effects of body size and temperature on the in situ reproduction, a multiple regression analysis was first run for egg production, body size and temperature. As the model showed a contribution of body size rather than temperature (see Results), we then used the weight-specific egg production in forward stepwise multiple regression analysis with the environmental variables, without correcting the egg production for temperature. Finally, we combined the significant factors from both models for a final model. For in situ reproduction, temperature and salinity were included as averaged values, while chl a and phytoplankton concentration/composition were included as integrated values for the 0–50 m water column. For fatty acids, only values from the chlorophyll maximum layer were available. As Acartia spp. in the North Sea appears to be relatively evenly distributed in the upper 40 m of the water column (P. Munk, DTU Aqua, Personal communication) we consider this to be a reasonable assumption.

The environmental variables used in the multiple regression model were chosen after data reduction by principal component analysis (PCA). The PCA was run on a correlation matrix with Varimax rotation, first separately for all the fatty acids, and then after combining the fatty acids which had the highest loading on different PCs with the other environmental variables. The PCA for fatty acids included 22 stations. In short, the PCA grouped the fatty acids according to the size fractions rather than type, so that most fatty acids in the size fraction 0–20 µm had a moderate or significant loading on PC3, most fatty acids in 20–40 µm had a significant loading on PC2 and most fatty acids in 40–80 µm had a significant loading on PC1 (data not shown). However, the percentage of PUFA from total fatty acids in 0–20 and 20–40 µm size fractions, and EPA:DHA ratio in the 40–80 µm size fraction, typically had a high loading on a different PC than the absolute concentrations. We thus chose the sum of PUFAs in different size classes (0–20, 20–40 and 40–80 µm), percentage of PUFAs in 0–20 and 20–40 µm size class and DHA:EPA ratio in 40–80 µm size class for further analysis with the other environmental factors.

Due to the missing data on phytoplankton biomass, the final PCA only included 11 stations. Roughly, the PCA grouped together diatoms with the total (autotrophic) microplankton biomass and EPA:DHA in 40–80 µm size fraction (PC1), heterotrophic biomass of all kinds (PC2), dinoflagellates and chl a (PC3), flagellates represented by a high PUFA percentage in the 0–20 µm fraction (PC4) and temperature with a high percentage of PUFA in 40–80 µm fraction (PC5; Table IIA). Based on this, we chose EPA:DHA in the 40–80 µm size fraction as a representative of PC1, ciliate biomass as a representative of PC2, chl a as a representative of PC3, percentage of PUFA in the 0–20 µm size fraction as a representative of PC4 and temperature as a representative of PC5, to be used together with the egg production and hatching in the multiple regression model. Chl a was used as a representative of PC3 rather than dinoflagellate biomass, as this allowed more stations (22) to be included in the model.

Table II:

Rotated component matrix (PCA with Varimax rotation) showing the loadings of different environmental variables in (A) 2001 and (B) 2003

 (A) Components in 2001
 
(B) Components in 2003
 
 PC 1 (27%) PC 2 (23%) PC 3 (16%) PC 4 (15%) PC 5 (11%) PC 1 (80%) PC 2 (16%) 
0.0054 −0.108 0.222 0.514 0.784   
Chl −0.519 −0.445 0.561 0.391 0.166 0.557 0.721 
−0.492 0.518 0.482 0.254 −0.183   
Flag 0.219 −0.376 −0.151 0.835 −0.111 0.867 0.443 
Het dino 0.336 −0.029 0.892 −0.146 −0.017 0.929 0.348 
Aut dino −0.218 −0.101 0.912 0.069 0.121 0.970 0.223 
Dino 0.080 −0.069 0.988 −0.048 0.054 0.962 0.253 
Diat 0.983 −0.011 −0.048 −0.071 −0.005 0.031 0.995 
Cil >50 µm 0.132 0.810 −0.116 −0.003 0.493   
Cil <50 µm −0.130 0.890 −0.265 −0.045 −0.255   
Cil tot −0.082 0.939 −0.252 −0.039 −0.108 0.375 0.905 
Aut tot 0.973 −0.097 −0.017 0.115 −0.020 0.965 0.248 
Het tot 0.137 0.886 0.332 −0.131 −0.115 0.664 0.742 
Micropl tot 0.978 0.004 0.020 0.099 −0.033 0.886 0.461 
PUFA 0–20 µm 0.628 0.329 0.057 0.562 −0.125   
PUFA% 0–20 µm −0.044 0.142 −0.038 0.928 −0.154   
PUFA 20–40 µm 0.306 −0.266 −0.080 −0.334 0.666   
PUFA% 20–40 µm −0.339 −0.396 0.136 0.641 0.369   
PUFA 40–80 µm −0.256 0.021 0.081 −0.206 0.878   
EPA:DHA 40–80 µm 0.961 0.008 0.140 −0.145 0.027   
FA tot      0.931 0.201 
EPA:DHA tot      0.339 0.924 
 (A) Components in 2001
 
(B) Components in 2003
 
 PC 1 (27%) PC 2 (23%) PC 3 (16%) PC 4 (15%) PC 5 (11%) PC 1 (80%) PC 2 (16%) 
0.0054 −0.108 0.222 0.514 0.784   
Chl −0.519 −0.445 0.561 0.391 0.166 0.557 0.721 
−0.492 0.518 0.482 0.254 −0.183   
Flag 0.219 −0.376 −0.151 0.835 −0.111 0.867 0.443 
Het dino 0.336 −0.029 0.892 −0.146 −0.017 0.929 0.348 
Aut dino −0.218 −0.101 0.912 0.069 0.121 0.970 0.223 
Dino 0.080 −0.069 0.988 −0.048 0.054 0.962 0.253 
Diat 0.983 −0.011 −0.048 −0.071 −0.005 0.031 0.995 
Cil >50 µm 0.132 0.810 −0.116 −0.003 0.493   
Cil <50 µm −0.130 0.890 −0.265 −0.045 −0.255   
Cil tot −0.082 0.939 −0.252 −0.039 −0.108 0.375 0.905 
Aut tot 0.973 −0.097 −0.017 0.115 −0.020 0.965 0.248 
Het tot 0.137 0.886 0.332 −0.131 −0.115 0.664 0.742 
Micropl tot 0.978 0.004 0.020 0.099 −0.033 0.886 0.461 
PUFA 0–20 µm 0.628 0.329 0.057 0.562 −0.125   
PUFA% 0–20 µm −0.044 0.142 −0.038 0.928 −0.154   
PUFA 20–40 µm 0.306 −0.266 −0.080 −0.334 0.666   
PUFA% 20–40 µm −0.339 −0.396 0.136 0.641 0.369   
PUFA 40–80 µm −0.256 0.021 0.081 −0.206 0.878   
EPA:DHA 40–80 µm 0.961 0.008 0.140 −0.145 0.027   
FA tot      0.931 0.201 
EPA:DHA tot      0.339 0.924 

The significant high loadings (>0.7) for each principal component (PC) are highlighted in bold, the significant moderate loadings (>0.5) are highlighted in italics; the variables which were chosen for the multiple regression analysis are underlined. (T) Temperature, (S) salinity, (Chl) chlorophyll a, (Flag) flagellates, (aut dino) autotrophic dinoflagellates, (het dino) heterotrophic dinoflagellates, (dino) total dinoflagellates, (diat) diatoms, (cil) ciliates, (aut tot) total autotrophs, (het tot) total heterotrophs, (micropl tot) total microplankton, (PUFA) polyunsaturated fatty acid concentration, (PUFA%) percentage of PUFAs from total fatty acids. All microplankton variables are biomass in carbon (µg L−1).

As standard copepods from a culture were used in the bioassay experiment, we did not need to account for the effect of body size or temperature, and only a single multiple regression was run. For all analysis, we used responses from Day 4 only, as this allowed a long adaptation time for the animals combined with simultaneous measurements of all environmental variables; the total number of observations was thus 8 (four stations with two water layers). Similar to the in situ reproduction, PCA was used to select for the environmental variables to be included in the model. All fatty acids had the highest loading in PC1, while only the ratio of EPA:DHA was separated with the highest loading on PC2 (data not shown). We thus included total fatty acids and the EPA:DHA ratio for the second PCA. The PCA combining fatty acids with other environmental variables separated flagellates and dinoflagellates together with a high concentration of total fatty acids (PC1) from diatoms and ciliates with a high EPA:DHA ratio (Table IIB). We thus chose total fatty acids as a representative of PC1 and EPA:DHA ratio as a representative of PC2. All statistics were conducted using the SigmaStat 3.1 or SPSS 11.5 statistical packages.

RESULTS

Environment

Seasonal variation at transect stations

The water column was well mixed in March and April, while in May and July a thermocline, accompanied with increased salinity, formed below ca. 20 m. There was no strong subsurface chlorophyll maximum layer, but chl a was either equally distributed in the upper 40 m (March and July), increasing slightly with depth (April), or clearly highest in the surface layer (May) (Fig. 3). The average water column temperature (0–50 m) over the year was rather similar at all stations, 4–6°C during March–April, close to 10°C in May and 11–15°C during July–September (Table III). The highest chl-a concentrations of ca. 5 µg L−1 (ca. 140 mg m−2) were observed during spring in March at T3 and May at T5, at other times and transects the average chl-a concentrations ranged between 1.6 and 3.5 µg L−1 (38–102 mg m−2). Diatoms dominated the microplankton composition at all stations and times and the proportion of heterotrophic organisms was typically between 10 and 20% of the total microplankton carbon. Polyunsaturated fatty acids (PUFAs) were abundant in the layer of maximum chl a, with concentrations up to ca. 50 µg L−1, forming between 14 and 36% of the total fatty acids (Table III).

Table III:

Average temperature, salinity (‰) and integrated chl-a concentration (mg m−2) for the 0–50 m water column (± SD), list of dominant algae groups with the number of stations in which the group dominates in parenthesis, integrated concentration of heterotrophic and total microplankton biomass (μg C L−1) with the percentage of heterotrophs in parenthesis, concentration of polyunsaturated fatty acids ARA (18:3n3), DHA (22:6n3), EPA (20:5n3) and total PUFA (μg L−1; with the percentage of PUFA from total FA in parenthesis), as well as the DHA:EPA ratio (μg:μg) in the chlorophyll maximum layer averaged over the different transects in March–September (T1–T5) and August (Bioassay study; S: surface water, M: chl max. layer)

Stations T, °C S (‰) Chl a, μg L−1 Group Heterotrophs, μg C L−1 (%) Microplankton, μg C L−1 ALA, μg L−1 DHA, μg L−1 EPA, μg L−1 DHA:EPA PUFA, μg L−1 (%) 
In situ 
 March T3 3.9 ± 1.1 33.7 ± 1.8 138 ± 121 Dia (1) 5.9 (5) 120 0.2 ± 0.02 0.1 ± 0.1 0.2 ± 0.1 1.7 ± 1.8 2.1 ± 0.9 (13.5) 
 April T1 4.7 ± 0.1 34.1 ± 0.5 63 ± 34 – 9.6 ± 1.7a – 0.9 ± 0.1 2.4 ± 0.2 1.1 ± 0.3 0.5 ± 0.2 13 ± 0.5 (16) 
 T3 5.4 34.7 78 – 3.7a – 3.1 1.1 3.5 0.3 21 (18) 
 T5 6.3 ± 0.1 32.9 ± 0.9 38 ± 16 Dia (3), cil (1) 40 ± 15 (17 ± 10) 221 ± 96 1.8 ± 1.6 5.8 ± 6.4 7.4 ± 7.2 0.7 ± 0.1 40 ± 28 (26) 
 May/June T5 9.7 ± 1.1 33.6 ± 0.5 144 ± 42 Dia (1), flag (1), dino (3) 26 ± 14 (20 ± 15) 171 ± 202 9.1 ± 14 5.6 ± 1.4 6.6 ± 5.4 1.1 ± 0.5 48 ± 33 (36) 
 July/August T1 14.0 ± 1.4 33.6 ± 0.3 85 ± 25 – 4.1 ± 1.8a – 0.7 ± 0.4 5.5 ± 3.9 4.4 ± 2.8 1.4 ± 0.6 17 ± 8 (34) 
 T3 11.3 ± 0.7 34.3 ± 0.1 102 ± 13 – 1.2 ± 1.5a – 1.1 ± 0.6 3.2 ± 1.4 2.1 ± 0.5 1.5 ± 0.3 14 ± 3 (21) 
 T4 11.8 ± 1.7 34.8 ± 0.3 65 ± 19 – 2.6a – 2.1 ± 2.0 4.4 ± 3.8 2.4 ± 1.3 0.9 ± 0.4 22 ± 18 (17) 
 T5 14.4 ± 1.0 33.4 ± 0.6 91 ± 3.3 Dia (1) 12.7 (14) 76 1.3 ± 1.9 2.1 ± 1.7 2.7 ± 3.4 0.6 ± 0.1 12 ± 14 (28) 
 September T1 15.3 ± 2.7 – – Dia (1), dino (1) 13 (20) 37 ± 4 – – – – – 
 T5 12.2 ± 5.2 – – Dia (1) 4.7 ± 4.0 (13 ± 12) 63 – – – – – 
Bioassay 
 S   0.3 ± 0.04 Flag + cil (4) 2.4 ± 1.3 (44 ± 13) 5.2 ± 2.5 0.2 ± 0.1 0.04 ± 0.1 0.01 ± 0.02 1.8 ± 2.5 3.3 ± 3.3 (5) 
 M   1.9 ± 0.9 Cil (2), dia (1), dino (1) 21 ± 12 (54 ± 25) 67 ± 56 1.1 ± 1.2 0.9 ± 0.8 0.1 ± 0.07 6.2 ± 3.4 12 ± 14 (12) 
Stations T, °C S (‰) Chl a, μg L−1 Group Heterotrophs, μg C L−1 (%) Microplankton, μg C L−1 ALA, μg L−1 DHA, μg L−1 EPA, μg L−1 DHA:EPA PUFA, μg L−1 (%) 
In situ 
 March T3 3.9 ± 1.1 33.7 ± 1.8 138 ± 121 Dia (1) 5.9 (5) 120 0.2 ± 0.02 0.1 ± 0.1 0.2 ± 0.1 1.7 ± 1.8 2.1 ± 0.9 (13.5) 
 April T1 4.7 ± 0.1 34.1 ± 0.5 63 ± 34 – 9.6 ± 1.7a – 0.9 ± 0.1 2.4 ± 0.2 1.1 ± 0.3 0.5 ± 0.2 13 ± 0.5 (16) 
 T3 5.4 34.7 78 – 3.7a – 3.1 1.1 3.5 0.3 21 (18) 
 T5 6.3 ± 0.1 32.9 ± 0.9 38 ± 16 Dia (3), cil (1) 40 ± 15 (17 ± 10) 221 ± 96 1.8 ± 1.6 5.8 ± 6.4 7.4 ± 7.2 0.7 ± 0.1 40 ± 28 (26) 
 May/June T5 9.7 ± 1.1 33.6 ± 0.5 144 ± 42 Dia (1), flag (1), dino (3) 26 ± 14 (20 ± 15) 171 ± 202 9.1 ± 14 5.6 ± 1.4 6.6 ± 5.4 1.1 ± 0.5 48 ± 33 (36) 
 July/August T1 14.0 ± 1.4 33.6 ± 0.3 85 ± 25 – 4.1 ± 1.8a – 0.7 ± 0.4 5.5 ± 3.9 4.4 ± 2.8 1.4 ± 0.6 17 ± 8 (34) 
 T3 11.3 ± 0.7 34.3 ± 0.1 102 ± 13 – 1.2 ± 1.5a – 1.1 ± 0.6 3.2 ± 1.4 2.1 ± 0.5 1.5 ± 0.3 14 ± 3 (21) 
 T4 11.8 ± 1.7 34.8 ± 0.3 65 ± 19 – 2.6a – 2.1 ± 2.0 4.4 ± 3.8 2.4 ± 1.3 0.9 ± 0.4 22 ± 18 (17) 
 T5 14.4 ± 1.0 33.4 ± 0.6 91 ± 3.3 Dia (1) 12.7 (14) 76 1.3 ± 1.9 2.1 ± 1.7 2.7 ± 3.4 0.6 ± 0.1 12 ± 14 (28) 
 September T1 15.3 ± 2.7 – – Dia (1), dino (1) 13 (20) 37 ± 4 – – – – – 
 T5 12.2 ± 5.2 – – Dia (1) 4.7 ± 4.0 (13 ± 12) 63 – – – – – 
Bioassay 
 S   0.3 ± 0.04 Flag + cil (4) 2.4 ± 1.3 (44 ± 13) 5.2 ± 2.5 0.2 ± 0.1 0.04 ± 0.1 0.01 ± 0.02 1.8 ± 2.5 3.3 ± 3.3 (5) 
 M   1.9 ± 0.9 Cil (2), dia (1), dino (1) 21 ± 12 (54 ± 25) 67 ± 56 1.1 ± 1.2 0.9 ± 0.8 0.1 ± 0.07 6.2 ± 3.4 12 ± 14 (12) 

For the location of transects and the number of stations per transect, see Fig. 1. and Table I (-) Missing data. Other abbreviations as in Table II.

aOnly ciliates included.

Fig. 3.

Example of a typical vertical distribution of temperature (°C; open circles), salinity (‰; solid triangles) and chl a (µg L−1; solid circles) in March (T3), April (T5), May (T5) and July (T5). For location of the transects, see Fig. 1.

Fig. 3.

Example of a typical vertical distribution of temperature (°C; open circles), salinity (‰; solid triangles) and chl a (µg L−1; solid circles) in March (T3), April (T5), May (T5) and July (T5). For location of the transects, see Fig. 1.

Year-to-year variation at the Dogger Bank

In all years, surface temperatures were high (ca. 18°C) and a thermocline was situated between ∼25 and 30 m (Fig. 4). Clear chlorophyll subsurface maxima were found just below the thermocline, with peak concentrations of 2–2.5 µg chl a L−1. Also the biomass of heterotrophic microplankton was higher in the chlorophyll maximum layers than at the surface, up to ca. 30 µg C L−1 (Fig. 4). During the bioassay study when microplankton composition and fatty acid concentration of seston were measured in more detail, about half of the microplankton was formed by heterotrophic dinoflagellates and ciliates (Table III), with an approximately equal contribution of autotrophic flagellates, dinoflagellates and diatoms. Both the absolute (≤12 µg L−1) and proportional (5–12%) concentrations of PUFA were lower during the bioassay study than on any of the transects sampled for in situ reproduction (Table III). Details of the environmental conditions in the sampling area in 2002 are presented by Poulsen and Kiørboe (Poulsen and Kiørboe, 2006), and in 2005 by Koski et al. (Koski et al., 2007) and Jónasdóttir and Koski (Jónasdóttir and Koski, submitted for publication).

Fig. 4.

Vertical distribution of temperature (°C; lines) and chl a (µg L−1; symbols) at Dogger Bank in 2002, 2003 and 2005. (Squares and solid line) station 2, (diamonds and medium dashed line) station 4, (downward triangles and long dashed line) station 1, (circles and dotted line) station 3, (upward triangles and solid line) station 5 and (stars and dash-dot line) station 8. The numbers indicate concentration of microzooplankton and heterotrophic flagellates (µg C L−1 ±SD) in surface and in chlorophyll maximum layer. For locations of the stations, see Fig. 1.

Fig. 4.

Vertical distribution of temperature (°C; lines) and chl a (µg L−1; symbols) at Dogger Bank in 2002, 2003 and 2005. (Squares and solid line) station 2, (diamonds and medium dashed line) station 4, (downward triangles and long dashed line) station 1, (circles and dotted line) station 3, (upward triangles and solid line) station 5 and (stars and dash-dot line) station 8. The numbers indicate concentration of microzooplankton and heterotrophic flagellates (µg C L−1 ±SD) in surface and in chlorophyll maximum layer. For locations of the stations, see Fig. 1.

Vertical distribution and biomass

No diel vertical migration was observed for any of the copepod species, but the different species seemed to consistently inhabit different depth layers (Fig. 5). While Pseudocalanus sp. and Temora longicornis seemed to primarily reside below the thermocline, copepod nauplii, Paracalanus parvus, Microcalanus pusillus and Oithona spp. were mainly found above the thermocline. Centropages typicus was primarily found above the thermocline in 2002, but in 2005 this species seemed to have a bimodal daytime distribution, with about half of the individuals residing below the thermocline. If the vertical distribution was separated into life stages, it appeared that all life stages (separated to nauplii, copepodites, males and females) of Oithona spp. and P. parvus preferred the surface layers. The distribution of other species was more spread with most of the copepodites of M. pusillus and C. typicus rather evenly distributed in the upper 40 m, and most of the copepodites of T. longicornis and Pseudocalanus sp. evenly distributed below 20 m. In contrast, adults of these species were most abundant between 30 and 40 m, thus close to the chlorophyll maximum layer, with the exception of Pseudocalanus sp. females, which were most abundant close to the bottom (data not shown).

Fig. 5.

Vertical distribution of copepod nauplii, Pseudocalanus sp., Centropages typicus, Paracalanus parvus, Microcalanus pusillus, Temora longicornis and Oithona spp. at the Dogger Bank in (A) 2002, (B) 2003 and (C) 2005 (ind. m−3; average of all stations ±SE). The shaded area indicates the chlorophyll maximum layer, the number indicates the proportion of the species from total biomass of small copepods (average of all stations ±SE). In 2002, black columns represent night samples (average of four sampling times), light grey columns day samples (average of five sampling times). Note different scales of the x-axis.

Fig. 5.

Vertical distribution of copepod nauplii, Pseudocalanus sp., Centropages typicus, Paracalanus parvus, Microcalanus pusillus, Temora longicornis and Oithona spp. at the Dogger Bank in (A) 2002, (B) 2003 and (C) 2005 (ind. m−3; average of all stations ±SE). The shaded area indicates the chlorophyll maximum layer, the number indicates the proportion of the species from total biomass of small copepods (average of all stations ±SE). In 2002, black columns represent night samples (average of four sampling times), light grey columns day samples (average of five sampling times). Note different scales of the x-axis.

If the vertical distribution was compared with the temperature and chl-a concentration, it appeared that Oithona spp., Paracalanus parvus and Centropages typicus were most abundant in the layer where temperature was >14°C, while Temora longicornis and Microcalanus pusillus were spread in temperatures between 6 and 15°C, and 7 and 18°C, respectively (Fig. 6A). Pseudocalanus spp. had a bimodal distribution with a high proportion of individuals residing either in rather cold (6–8°C) or rather warm (>14°C) water layers. With the possible exception of adult stages of T. longicornis, M. pusillus and C. typicus, none of the species appeared to be strongly associated with the chlorophyll maximum layer, as the peak distribution of most life stages of all species typically was at ≤1 µg Chl a L−1 (Fig. 6B). Similarly, no effect of salinity was observed for any of the species or life stages (data not shown).

Fig. 6.

The proportional distribution of different life stages (% from total) as a function of (A) temperature, and (B) chl a. (Open symbols with a cross) nauplii, (open symbols) copepodites I–V, (solid symbols) females, (grey symbols) males. (Triangles) 2002, (circles) 2003, (squares) 2005. Only species and years representing ≤9% of the total biomass of small copepods are included in the figure (Fig. 5).

Fig. 6.

The proportional distribution of different life stages (% from total) as a function of (A) temperature, and (B) chl a. (Open symbols with a cross) nauplii, (open symbols) copepodites I–V, (solid symbols) females, (grey symbols) males. (Triangles) 2002, (circles) 2003, (squares) 2005. Only species and years representing ≤9% of the total biomass of small copepods are included in the figure (Fig. 5).

The depth-integrated biomass of small copepods varied threefold, with the lowest values observed in 2002 (90–120 mg C m−2), while in 2003 and 2005 the biomass ranged from 130 to 290 mg C m−2. Typically, sac-spawners (Oithona spp.) dominated the biomass in 2002, while in 2003 all stations were dominated by calanoid broadcast-spawning species (Microcalanus pusillus and Centropages typicus), and in 2005 the dominance switched between broadcast spawners (mainly Temora longicornis; station 8) and sac-spawners (Oithona spp.; station 5; Table IV, Fig. 5). No significant correlation was observed between the depth-integrated copepod biomass and phytoplankton (average and max. chl-a concentration) or microzooplankton concentration, neither between the total biomass and temperature (Pearson P > 0.05; data not shown). However, the biomass of sac-spawning copepods was significantly negatively correlated with temperature (Pearson correlation coefficient −0.789; P < 0.05).

Table IV:

Depth-integrated biomass (Bm; mg C m−2) and secondary production (Sp; mg C m−2 day−1) of broadcast and sac-spawning copepods, as well as the average egg production (Ep; % body weight day−1) used to calculate the secondary production, and the species dominating the biomass of small copepods (Acartia spp., Temora longicornis, Pseudocalanus sp., Centropages typicus, Microcalanus pusillus, Paracalanus parvus and Oithona spp.) in 2002, 2003 and 2005 (Fig. 5)

 Broadcast spawners
 
Sac spawners
 
Total
 
Station Bm Ep Sp Bm Ep Sp Sp Dominant species 
2002 
 St. 2 6.4 10.0 0.6 75.2 2.1 12.1 12.8 Oithona spp. 
 St. 4 18.7 10.0 1.9 37.8 2.4 6.1 Oithona spp., C. typicus 
2003 
 St. 3 97.1 11.0 11.9 23.8 5.0 1.0 12.9 M. pusillus, C. typicus 
 St. 5 204.8 8.1 17.9 22.4 6.0 1.3 19.2 C. typicus, M. pusillus 
 St. 8 165.4 9.4 17.6 19.7 4.8 0.9 18.6 M. pusillus 
2005 
 St. 1 114.3 10.2 11.7 115.5 2.1 2.4 14.1 Oithona spp., T. longicornis 
 St. 3 167.0 4.4 7.3 117.5 1.0 1.2 8.5 T. longicornis, Oithona spp. 
 St. 5 39.6 13.8 4.2 134.8 0.5 0.7 4.9 Oithona spp., Pseudocalanus sp. 
 St. 8 168.0 9.8 16.5 57.7 2.0 1.1 17.7 T. longicornis, Paracalanus parvus, Oithona spp. 
 Broadcast spawners
 
Sac spawners
 
Total
 
Station Bm Ep Sp Bm Ep Sp Sp Dominant species 
2002 
 St. 2 6.4 10.0 0.6 75.2 2.1 12.1 12.8 Oithona spp. 
 St. 4 18.7 10.0 1.9 37.8 2.4 6.1 Oithona spp., C. typicus 
2003 
 St. 3 97.1 11.0 11.9 23.8 5.0 1.0 12.9 M. pusillus, C. typicus 
 St. 5 204.8 8.1 17.9 22.4 6.0 1.3 19.2 C. typicus, M. pusillus 
 St. 8 165.4 9.4 17.6 19.7 4.8 0.9 18.6 M. pusillus 
2005 
 St. 1 114.3 10.2 11.7 115.5 2.1 2.4 14.1 Oithona spp., T. longicornis 
 St. 3 167.0 4.4 7.3 117.5 1.0 1.2 8.5 T. longicornis, Oithona spp. 
 St. 5 39.6 13.8 4.2 134.8 0.5 0.7 4.9 Oithona spp., Pseudocalanus sp. 
 St. 8 168.0 9.8 16.5 57.7 2.0 1.1 17.7 T. longicornis, Paracalanus parvus, Oithona spp. 

The egg production of sac-spawning copepods was calculated using the egg ratio method; the egg production of broadcast spawners was based on the significant regression between prosome length and egg production (Fig. 2; Methods). For station localities, see Fig. 1.

Egg production and hatching

In situ reproduction

The egg production rate of Acartia clausi was highest in April (21 ± 15 eggs f−1 day−1; mean ± SD), due to the significantly higher number of eggs produced per brood compared to the other months (Kruskal–Wallis; H3 = 95; P < 0.001). However, there were also significant differences in egg production between stations (one-way ANOVA, F6 = 6.7, P < 0.001 for April; Kruskal–Wallis, H4 = 45.5; P < 0.001 for June and Kruskal–Wallis, H7 = 64.8, P < 0.001 for July), sometimes with >10-fold differences in egg production even between the stations on the same transect (Fig. 7A). Hatching success over the year varied between 60 and 75%, but due to the short incubation time, the spring hatching results had to be omitted. The between-station variation in hatching success was typically less than the variation in egg production, although a few exceptionally low hatching successes of 26 and 33% were observed at stations 5.7 in June and 3.4 in July, respectively. The female prosome length indicated that the spring, early summer and late summer individuals belonged to different generations (Kruskal–Wallis; H3 = 252; P < 0.001; Table V). Due to the low female abundance in spring, the high egg production in April did not result in a high number of nauplii, but the nauplii peak was observed in late summer during the peak abundance of females (Fig. 7A).

Table V:

Average egg production (eggs f−1 day−1), percentage of females producing eggs (% day−1) and brood size (no of eggs brood−1), hatching success (%) and female prosome length (mm) of Acartia clausi and Centropages typicus in March–September (in situ reproduction), and Acartia tonsa in August 2003 (bioassay reproduction; mean ± SD)

 Egg production
 
Hatching Length 
 Eggs f−1 day−1 % prod Eggs brood−1 mm 
(A) A. clausi 
March (1) 8.2 ± 7.0a,b 67a 11 ± 6b – 0.9 ± 0.1b,c 
April (7) 21 ± 15a 82 ± 17a 25 ± 6a – 1.1 ± 0.07a 
May–June (5) 8.1 ± 7.2b 71 ± 36a 9 ± 5b 60 ± 22 (320)a 1.0 ± 0.08b 
July–August (9) 6.1 ± 6.9b 52 ± 33a 9 ± 4b 75 ± 18 (205)a 0.9 ± 0.06c 
(B) C. typicus 
March (1) 34 ± 24a 82a 42 ± 19a – 1.35 ± 0.05a 
July–August (5) 65 ± 56a 75 ± 34a 76 ± 13a 36 ± 11 (2643) 1.27 ± 0.07b 
September (4) 14 ± 17b 63 ± 36a 19 ± 7b – 1.26 ± 0.05b 
August 2003 (1) 39 ± 29 87 45 ± 26 – 1.20 ± 0.04 
(C) A. tonsa 
Chl max (4) 4.7 ± 4.2a   81 ± 8 (534)a  
Surface (4) 1.4 ± 1.3b   86 ± 13 (158)a  
 Egg production
 
Hatching Length 
 Eggs f−1 day−1 % prod Eggs brood−1 mm 
(A) A. clausi 
March (1) 8.2 ± 7.0a,b 67a 11 ± 6b – 0.9 ± 0.1b,c 
April (7) 21 ± 15a 82 ± 17a 25 ± 6a – 1.1 ± 0.07a 
May–June (5) 8.1 ± 7.2b 71 ± 36a 9 ± 5b 60 ± 22 (320)a 1.0 ± 0.08b 
July–August (9) 6.1 ± 6.9b 52 ± 33a 9 ± 4b 75 ± 18 (205)a 0.9 ± 0.06c 
(B) C. typicus 
March (1) 34 ± 24a 82a 42 ± 19a – 1.35 ± 0.05a 
July–August (5) 65 ± 56a 75 ± 34a 76 ± 13a 36 ± 11 (2643) 1.27 ± 0.07b 
September (4) 14 ± 17b 63 ± 36a 19 ± 7b – 1.26 ± 0.05b 
August 2003 (1) 39 ± 29 87 45 ± 26 – 1.20 ± 0.04 
(C) A. tonsa 
Chl max (4) 4.7 ± 4.2a   81 ± 8 (534)a  
Surface (4) 1.4 ± 1.3b   86 ± 13 (158)a  

The number in parenthesis after the month indicates the number of stations, the number after hatching success the number of observed eggs. The letters group the months which are not significantly different from each other (P > 0.05). (Blank) no experiment, (–) missing data.

Fig. 7.

Female abundance (ind. m−2), in situ egg production (eggs f−1 day−1; mean ± SE) and nauplii abundance (ind. m−2) of (A) Acartia clausi and (B) Centropages typicus along the North Sea transects in March–September. The transect/station numbers as in Fig. 1 and Table III. The calculated nauplii abundance refers to the amount of nauplii estimated based on female abundance, egg production and hatching success, while the observed abundance is the number of nauplii in zooplankton (pump) samples (see Methods). (MD) Missing data.

Fig. 7.

Female abundance (ind. m−2), in situ egg production (eggs f−1 day−1; mean ± SE) and nauplii abundance (ind. m−2) of (A) Acartia clausi and (B) Centropages typicus along the North Sea transects in March–September. The transect/station numbers as in Fig. 1 and Table III. The calculated nauplii abundance refers to the amount of nauplii estimated based on female abundance, egg production and hatching success, while the observed abundance is the number of nauplii in zooplankton (pump) samples (see Methods). (MD) Missing data.

Centropages typicus egg production was highest in late summer (65 ± 56 eggs f−1 day−1), intermediate in spring (34 ± 24 eggs f−1 day−1) and low in autumn (14 ± 17 eggs f−1 day−1), due to the significantly lower number of eggs per brood in September than in March or July (Kruskal–Wallis; H2 = 70; P < 0.001). Similar to Acartia clausi, also C. typicus egg production was significantly different between stations, both in July (Kruskal–Wallis; H4 = 17.0; P < 0.01) and in September (Kruskal–Wallis; H3 = 29.8; P < 0.001), although with typically lower between-station variation (Fig. 7B). When measured, the hatching success was relatively low (36 ± 11%). Based on the female prosome length, it appeared that the spring individuals belonged to a different generation than summer and autumn individuals (Kruskal–Wallis; H2 = 24; P < 0.001; Table V). Similar to Acartia clausi, the high egg production of spring and summer did not result in a nauplii peak which only occurred during the high female abundance in September (Fig. 7B).

The egg production rate of Acartia clausi was significantly positively correlated with prosome length, and significantly negatively correlated with temperature (Pearson correlation; P < 0.01; Fig. 8), while hatching did not correlate with prosome length, temperature or egg production (Pearson; P > 0.5; data not shown). Egg production of C. typicus was not significantly related to body size or temperature (Pearson; P > 0.05; Fig. 8). As a multiple correlation analysis with egg production of A. clausi against body size and temperature showed that the variation in egg production was mostly explained by body size (Table VIA), we decided to use the weight-specific egg production in the multiple regression analysis with food variables, but not to correct the egg production for temperature. Based on the PCA, we chose one environmental variable representing each PC for the multiple correlation analysis with the weight-specific egg production and hatching success. The analysis showed that PC2 and PC4 represented by EPA:DHA ratio in the size class of 40–80 µm and percentage of PUFA from the total fatty acids in the 0–40 µm size fraction had a significant effect on egg production, while only the EPA:DHA ratio in the 40–80 µm size class had a significant effect on hatching success. The model implied an increase of 0.02 µg C (µg C)−1 day−1 in weight-specific egg production and a decrease of 19% in hatching success with a unit change in EPA:DHA ratio and a decrease of 0.001 µg C (µg C)−1 day−1 with a unit change in the percentage of 0–40 µm PUFAs (Table VIA). Along with the EPA:DHA ratio in 40–80 µm, also diatoms, total autotrophs and total microplankton concentration had a significant loading on PC2, while flagellates had a significant loading along with the PUFA proportion in PC4 (Table IIA), so that the effect of these factors could not be separated.

Table VI:

Multiple regression coefficients and their significance for the models describing the factors influencing (A) the in situ egg production and hatching of Acartia clausi and (B) the bioassay egg production and hatching of Acartia tonsa

   Coefficient SE F-to-remove F-to-enter P-value 
A. Egg production Constant −37.31 32.6    
 N = 22 C1 (body size) 56.18 28.9   0.067 
 R2 = 0.487 C2 (temperature) −0.634 0.575   0.284 
 Adj. R2 = 0.434       
 Egg production Constant 0.119 0.033    
 N = 18 C1 (PUFA% 0–40 µm) −0.001 0.0005 8.481  0.011* 
 R2 = 0.613 C2 (EPA:DHA 40–80 µm) 0.022 0.008 7.499  0.015* 
 Adj. R2 = 0.561 C3 (T)    1.267 0.278 
  C4 (Chl a   0.111 0.743 
  C5 (Ciliate bm)    1.978 0.18 
 Egg production Constant −39.53 18.9    
 N = 20 C1 (body size) 45.79 20.7   0.041* 
 R2 = 0.68 C2 (EPA:DHA 40–80 µm) 4.828 1.67   0.010* 
 Adj. R2 = 0.64       
 Hatching Constant 74.762 3.733    
 N = 11 C1 (PUFA% 0–40 µm)    0.248 0.63 
 R2 = 0.577 C2 (EPA:DHA 40–80 µm) −18.992 5.419 12.284  0.007** 
 Adj. R2 = 0.530 C3 (T)    0.083 0.779 
  C4 (Chl a   0.373 0.555 
  C5 (Ciliate bm)    1.573 0.238 
B. Egg production Constant 2.684 0.75    
 N = 8 C1 (EPA:DHA) 11.396 5.175 4.849  0.07 
 R2 = 0.447 C2 (Total FA)    0.118 0.743 
 Adj. R2 = 0.355       
 Hatching Constant 93.673 4.285    
 N = 7 C1 (EPA:DHA) −59.720 24.144 6.188  0.056 
 R2 = 0.550 C2 (Total FA)    0.501 0.511 
 Adj. R2 = 0.460       
   Coefficient SE F-to-remove F-to-enter P-value 
A. Egg production Constant −37.31 32.6    
 N = 22 C1 (body size) 56.18 28.9   0.067 
 R2 = 0.487 C2 (temperature) −0.634 0.575   0.284 
 Adj. R2 = 0.434       
 Egg production Constant 0.119 0.033    
 N = 18 C1 (PUFA% 0–40 µm) −0.001 0.0005 8.481  0.011* 
 R2 = 0.613 C2 (EPA:DHA 40–80 µm) 0.022 0.008 7.499  0.015* 
 Adj. R2 = 0.561 C3 (T)    1.267 0.278 
  C4 (Chl a   0.111 0.743 
  C5 (Ciliate bm)    1.978 0.18 
 Egg production Constant −39.53 18.9    
 N = 20 C1 (body size) 45.79 20.7   0.041* 
 R2 = 0.68 C2 (EPA:DHA 40–80 µm) 4.828 1.67   0.010* 
 Adj. R2 = 0.64       
 Hatching Constant 74.762 3.733    
 N = 11 C1 (PUFA% 0–40 µm)    0.248 0.63 
 R2 = 0.577 C2 (EPA:DHA 40–80 µm) −18.992 5.419 12.284  0.007** 
 Adj. R2 = 0.530 C3 (T)    0.083 0.779 
  C4 (Chl a   0.373 0.555 
  C5 (Ciliate bm)    1.573 0.238 
B. Egg production Constant 2.684 0.75    
 N = 8 C1 (EPA:DHA) 11.396 5.175 4.849  0.07 
 R2 = 0.447 C2 (Total FA)    0.118 0.743 
 Adj. R2 = 0.355       
 Hatching Constant 93.673 4.285    
 N = 7 C1 (EPA:DHA) −59.720 24.144 6.188  0.056 
 R2 = 0.550 C2 (Total FA)    0.501 0.511 
 Adj. R2 = 0.460       

In (A) [Egg production = constant + (C1 body size) + (C2 temperature)], [Weight-specific egg production/hatching = constant + (C1 PUFA% in the 0–40 µm size fraction) + (C2 EPA:DHA in the 40–80 µm size fraction) + (C3 T) + (C4 Chl a) + (C5 ciliate biomass)] and [Egg production = constant + (C1 body size) + (C2 EPA:DHA in 40–80 µm)]. In (B) [Egg production/hatching = constant + (C1 EPA:DHA) + (C2 total fatty acid concentration)]. The values represent the best fit of a multiple linear regression for egg production, body size and temperature (A. clausi) and the best fit of a forward stepwise regression for the weight-specific egg production and environmental variables (A. clausi) and for all regressions with A. tonsa. Weight-specific egg production is expressed as µg C (µg C)−1 day−1, egg production in number of eggs f−1 day−1, body size as length in mm, chl a and total fatty acids as µg L−1, ciliate biomass as µg C L−1, PUFA as a percentage from total fatty acids and EPA:DHA as a weight ration (µg:µg). The coefficients describe the expected increase/decrease in egg production or hatching per a unit change of the independent variable. *Denotes the significant coefficients at the P < 0.05 level, **significant coefficients at the P < 0.01 level.

Fig. 8.

Egg production (eggs f−1 day−1; mean ± SE) of Acartia clausi and Centropages typicus as a function of prosome length (mm; mean ± SE) and temperature (°C). The significant correlation coefficients from the Pearson correlation analysis are shown in the figure.

Fig. 8.

Egg production (eggs f−1 day−1; mean ± SE) of Acartia clausi and Centropages typicus as a function of prosome length (mm; mean ± SE) and temperature (°C). The significant correlation coefficients from the Pearson correlation analysis are shown in the figure.

Both the model with body size and the model with food variables, however, only explained ≤60% of the variation in egg production. We thus decided to combine these two and ran the multiple regression analysis with the egg production, body size and EPA:DHA ratio in 40–80 µm size class. The combination of body size and EPA:DHA ratio in 40–80 µm (representing the biomass of diatoms and autotrophic microplankton in general) explained 68% of the variance in egg production, with an increase of 46 eggs f−1 day−1 with a millimetre increase in prosome length and an increase of 4.8 eggs f−1 day−1 with a unit increase in EPA:DHA ratio. Due to scarcity of data, a multiple regression analysis was not attempted for Centropages typicus egg production or hatching success.

Bioassay

Egg production of the cultured Acartia tonsa fed with in situ water was generally low, on average ca. 5 and 1.5 eggs f−1 day−1, in the chlorophyll maximum and surface layers, respectively (Fig. 9A, Table V). The cumulative egg production (Day 3) was significantly different between the stations (two-way ANOVA; F3,33 = 9.8; P < 0.001) and the water layers (F1,33 = 61.5; P < 0.001), with significantly highest egg production at station 8 and in the chlorophyll maximum layer (Tukey; P < 0.05). The average egg production in the chlorophyll maximum layer was similar to the egg production on Rhodomonas sp. (Tukey HSD; P > 0.05). Hatching success was high, respectively, 81 and 86% in the surface and chlorophyll maximum layers, and similar to the hatching on Rhodomonas sp. (Fig. 9B, Table V). The multiple regression analysis confirmed the positive effect of increasing EPA:DHA ratio (here associated with diatom and ciliate biomasses) on egg production and its negative effect on hatching success, as observed with the in situ reproduction. Model coefficients suggested an increase in egg production by 11.4 eggs f−1 day−1 and a decrease in hatching success by 59.7% with a unit change in EPA:DHA ratio. However, the model explained only 45% of the egg production and 55% of the hatching success of Acartia tonsa (Table VIB).

Fig. 9.

(A) Cumulative egg production during the five subsequent days of incubation (eggs f−1) and (B) average hatching success (%) of cultured Acartia tonsa incubated in water collected from chlorophyll maximum (solid symbols, black columns) and surface (open symbols, white columns) layers at the four Dogger Bank stations, as well as in Rhodomonas sp. (grey symbols and columns; mean ± SE). In (B), 1, 3, 5 and 8 refer to the stations numbers, M and S to chlorophyll maximum and surface layers, respectively.

Fig. 9.

(A) Cumulative egg production during the five subsequent days of incubation (eggs f−1) and (B) average hatching success (%) of cultured Acartia tonsa incubated in water collected from chlorophyll maximum (solid symbols, black columns) and surface (open symbols, white columns) layers at the four Dogger Bank stations, as well as in Rhodomonas sp. (grey symbols and columns; mean ± SE). In (B), 1, 3, 5 and 8 refer to the stations numbers, M and S to chlorophyll maximum and surface layers, respectively.

Secondary production

The secondary production of dominant neritic copepods in July–August was typically highest at the shallow station 8 (18–19 mg C m−2 day−1). The secondary production at this station was mostly formed by broadcast-spawning calanoid species. At most other stations, the secondary production ranged from 8 to 14 mg C m−2 day−1, with the exception of station 4/5 which in 2003 had a high secondary production of 19 mg C m−2 day−1, but in 2002 and 2005 a low secondary production of 5–8 mg C m−2 day−1. The high secondary production in 2003 was associated with a high biomass of broadcast-spawning copepods, while in 2002 and 2005 sac-spawning copepods dominated the biomass at this station (Table IV). Secondary production thus did not always correlate with the biomass, but a high proportion of broadcast-spawning species tended to lead to a high secondary production, likely due to their higher weight-specific egg production (Table IV). No significant correlations were observed between secondary production and phytoplankton (average or maximum chl a) or microzooplankton biomass (Pearson, P > 0.05; data not shown), but secondary production was significantly positively correlated with the average water column temperature (Fig. 10).

Fig. 10.

Secondary production (mg C m−2 day−1) of small copepods as a function of average water column temperature.

Fig. 10.

Secondary production (mg C m−2 day−1) of small copepods as a function of average water column temperature.

DISCUSSION

In summary, our results showed that the vertical distribution of small copepods was not determined by subsurface chlorophyll maximum, but only temperature seemed to have an effect on the distribution of (some) species. Secondary production, however, was highest at the most productive station 8 on top of the Dogger Bank, and egg production seemed to benefit from the chlorophyll maximum particularly at this station. Secondary production appeared to be governed mostly by the copepod species composition and reproductive rate, as the broadcast-spawning calanoid species had a substantially higher weight-specific reproduction than e.g. Oithona spp. While reproduction seemed to be controlled by the changes in body size and qualitative aspects of food (such as the EPA:DHA ratio), temperature appeared to have an influence on species composition (broadcast versus sac-spawning species). The seasonal differences in temperature (body size) and food quality likely overrode the potential effects of food quantity, which otherwise could have manifested in a high reproduction at productive frontal stations. These findings and their relevance are discussed in detail below.

Secondary production on Dogger Bank

It has been suggested that the frontal zones in the North Sea are hot spots of biological activity, where production, food-web structure and diversity change within short distances (Nielsen and Munk, 1998). In these areas, the elevated chlorophyll concentration would support high zooplankton production, with an associated biomass peak of copepods in the chlorophyll maximum layer (Nielsen et al., 1993; Richardson et al., 2000). On the Dogger Bank, stations 5 and 8 could be considered the most productive, as they were located closest to the area with the upwelling of new nutrients (Richardson et al., 2000). Our study confirmed the high secondary production at these stations, though at station 5 only when the small copepods were dominated by broadcast-spawning calanoids. The other potential frontal stations, 5.1, 5.2 and the stations in the transect 1 (Fig. 1) did not show a clear subsurface chlorophyll maximum, but the chl-a values were low throughout the summer. Consequently, reproduction was not elevated at these stations (Fig. 7), and reproduction on an annual scale appeared controlled by other factors than chl a.

The highest depth-integrated biomass and secondary production at stations 5 and 8 were similar to those which have been reported previously for the small copepods in the frontal areas of the North Sea, but lower than the typical biomass in the coastal areas where the small copepod species dominate (Nielsen and Munk, 1998). It appeared that the high biomass did not directly lead to a high secondary production (Table IV), but species composition (broadcast versus sac-spawners) and weight-specific reproduction determined the secondary production. Neither biomass nor secondary production was connected to chlorophyll concentration, but temperature seemed to have an effect on dominance of specific groups. It seemed that the biomass of sac-spawning copepods was negatively related to temperature, while the total secondary production increased with increasing temperature, with a highest depth-integrated production occurring at temperatures >13°C (Fig. 10). This was likely due to the high abundance of two out of three dominant calanoid broadcast-spawning species (Centropages typicus and Microcalanus pusillus) at the warm temperatures of ca. 18°C (Fig. 6), although temperature could also have a direct influence on the reproduction rate (e.g. Koski and Kuosa, 1999; Table VI). If the secondary production of small copepods was compared with that of Calanus spp. (Jónasdóttir and Koski, submitted for publication), it appeared that the small copepods dominated the secondary production at station 8, while at stations 3 and 5 further away from the Dogger Bank secondary production of Calanus spp. was approximately two to six times higher than that of small copepods. However, there was a large between-year variation in secondary production of both Calanus spp. and small copepods at these stations, with two to three times higher secondary production of small copepods in 2002, but two to six times higher secondary production of Calanus spp. in 2005 (Jónasdóttir and Koski, submitted for publication).

The effect of subsurface chlorophyll maximum on vertical distribution and reproduction

Our results for the vertical distribution of dominant small copepod species did not support the observations of highest biomass in the subsurface chlorophyll maximum (Fig. 5), but were more similar to the studies showing maximum abundances below or above the pycnocline (Harris, 1988; Hays et al., 1996). It appeared that only adult stages of some copepod species (Temora longicornis, Centropages typicus and Microcalanus pusillus) resided in or close to the chlorophyll maximum layer, while juvenile stages and all stages of other species preferred layers above or below the chlorophyll maximum.

In principal, the choice of the habitat layer of copepods should represent the best choice between maximum food abundance and minimum predation risk (Fossheim and Primicerio, 2008). Although visual predators (such as planktivorous fish) mainly forage in the surface layers, invertebrate predators can be abundant in the subsurface chlorophyll maximum (Fossheim and Primicerio, 2008). Some of the main invertebrate predators are chaetognaths, which have been shown to feed selectively on some small copepods, such as Pseudocalanus sp., and may control their biomass in late summer and autumn (Tönnesson and Tiselius, 2005). Although no significant connection between the abundance of different copepod species in the chlorophyll maximum and chaetognath abundance was observed in the study area (data not shown), it appeared that in 2003 the maximum abundance of chaetognaths (ca. 50 ind. m−3) was located close to the surface, while in 2002 and 2005 when copepods were typically either above or below the chlorophyll maximum layer, chaetognath abundances were at their maximum (20–50 ind. m−3) in the chlorophyll maximum. It could be that the vertical distribution of invertebrate predators contributes to the vertical distribution of small copepods, along with food availability, temperature and salinity. However, in the present study, we only observed weak suggestions of the effects of temperature and invertebrate predators and no effects of salinity or subsurface chlorophyll maximum.

Harris (Harris, 1988) suggested that, although copepods rarely reside in the subsurface chlorophyll maximum, they fill their guts and produce maximum amount of eggs while swimming through it during vertical migration. This behaviour was shown with Calanus helgolandicus, while in the study of Harris (Harris, 1988), Acartia spp. was the only copepod species which was permanently associated with the chlorophyll maximum layer. The subsurface chlorophyll maximum layer could thus be especially profitable for these species, although other neritic copepods would neither migrate through nor reside in the chlorophyll maximum layer (Fig. 5).

The egg production of Acartia tonsa was always higher when incubated in chlorophyll maximum water than in surface water at all stations. It has been suggested that the abundant dinoflagellates in the subsurface chlorophyll blooms would be profitable for copepod production (Richardson et al., 1998). However, this was not case in our study (Table VIA), although on average 55% of the phytoplankton biomass in the subsurface maximum was composed of dinoflagellates. The most abundant dinoflagellates during our study were various Ceratium species, which due to their large size and spiny morphology are typically not ingested by Acartia spp., although larger copepods such as Calanus spp. might graze on them (Jansen et al., 2006).

It has also been shown that phytoplankton in the subsurface chlorophyll maximum is less nutrient limited than phytoplankton in the surface layer (Richardson et al., 1998, 2000). As exponentially growing algae are generally of a better quality food than senescent ones (Jónasdóttir et al., 1995), a positive effect of food quality could be expected due to a higher nutrient concentrations in the chlorophyll maximum layer. If we assume that most of the phytoplankton diet of Acartia tonsa was formed by diatoms and flagellates (thus not Ceratium), and plot the egg production rate as a function of the concentration of these groups for the two layers, it indeed appears that the egg production was generally higher while feeding on chlorophyll maximum algae than on the corresponding concentration of surface algae (Fig. 11A). This apparent higher quality could have been related to the proportional PUFA content, which was higher in the chlorophyll maximum (12 ± 6%) than in the surface (5.3 ± 0.7), the high EPA:DHA ratio as suggested by the multiple regression analysis, or possibly to e.g. lower C:N ratio of the seston, which was not measured in the present study. Whatever the reason for improved quality, it did not affect the hatching, as no difference between the layers was observed for the hatching success (Fig. 11B).

Fig. 11.

(A) Egg production (eggs f−1 day−1) and (B) hatching success (%) of Acartia tonsa as a function of concentration of diatoms and flagellates (µg C L−1) in the bioassay experiment (mean ± SE). Only the range of concentrations occurring both in the surface and chlorophyll maximum layer are included. (Filled circles) Surface layer, (open circles) chlorophyll maximum layer.

Fig. 11.

(A) Egg production (eggs f−1 day−1) and (B) hatching success (%) of Acartia tonsa as a function of concentration of diatoms and flagellates (µg C L−1) in the bioassay experiment (mean ± SE). Only the range of concentrations occurring both in the surface and chlorophyll maximum layer are included. (Filled circles) Surface layer, (open circles) chlorophyll maximum layer.

Factors controlling reproduction

We observed both seasonal (between months) and spatial (between stations) variation in egg production. As observed earlier, Acartia clausi egg production was positively related to prosome length and negatively related to temperature, likely reflecting the negative effect of temperature on body size (e.g. Halsband and Hirche, 2001). As the body size or temperature rarely differed between the sampling stations (during the same month), the effect of body size on egg production was mainly relevant at an annual scale. If the effect of body size was removed by using the weight-specific egg production, the egg production appeared to be dependent on the qualitative aspects of the food, such as proportional PUFA content and EPA:DHA ratio. This effect of food could have been due to both spatial and seasonal patchiness in the food environment and could thus affect the egg production both at short and long (annual) time scales. The beneficial effect of a high EPA:DHA ratio was also suggested by the bioassay experiments.

EPA:DHA ratio has been observed to correlate with egg production in several studies, sometimes positively and sometimes negatively (reviewed by Jónasdóttir et al., 2009). The reason for the correlation between egg production and EPA:DHA ratio is unclear, although it is assumed that it has to do with the chemical similarity of these compounds and the following competition in chemical and physiological reactions (Sargent et al., 1999). However, in our study, the beneficial effect of a high EPA:DHA ratio and a high percentage of PUFA likely reflected a positive effect of autotrophic food sources in general (diatoms in particular), as a high EPA:DHA ratio also reflected high concentrations of diatoms and total autotrophic microplankton (Table II). This suggested that the body size (likely controlled by temperature) and specific fatty acids abundant in phytoplankton were the main factors controlling egg production. Our results are thus in agreement with previous field studies demonstrating a combined effect of temperature (via body size) and food (Calbet and Agustí, 1999; Bunker and Hirst, 2004; Maps et al., 2005), although EPA:DHA ratio (or diatom biomass) rather than bulk chl a appeared to be a better predictor of egg production.

As observed earlier, egg production and hatching success were controlled by different factors; in fact while the high EPA:DHA ratio was beneficial for egg production, it appeared to have a negative effect for hatching success. Typically, high hatching success correlates with a high ingestion of DHA, while a negative correlation is often observed between hatching success and EPA:DHA ratio (Jónasdóttir et al., 2009). High EPA concentration of seston is mostly associated with a high biomass of diatoms, which typically promotes high egg production but low hatching success (e.g. Peters et al., 2007). As diatoms often dominate the phytoplankton in the coastal areas of the North Sea, particularly during the spring bloom, a low hatching success, together with a low female abundance, can lead to low recruitment irrespective of the high egg production rate.

In the present study, the maximum egg production of Acartia clausi occurred in April, and was similar to the rate previously recorded for this species (Halsband and Hirche, 2001; Wesche et al., 2007). Due to the low number of females in the water column, the high egg production in the spring was, however, not followed by an increase in nauplii abundance, which only peaked in the summer. Similarly, the peak egg production of Centropages typicus was highest in July–August, but the peak abundance of nauplii first occurred 2 months later in September. Although the irregular sampling and advection did not allow a detailed analysis of population dynamics, it appeared clear that the high egg production in spring (Acartia clausi) or summer (Centropages typicus) did not lead to an increase in biomass. For A. clausi, it is also likely that the female peak in June did not originate from April eggs, but rather March eggs: even in optimal food conditions the generation time of A. clausi at the 6–7°C would be >60 days (Huntley and Lopez, 1992; Klein Breteler and Schogt, 1994), while the time between the sampling events was ca. 40 days. The April cohort thus more or less disappeared from the water column. It seems that the success of spring generations of small copepods can be extremely variable, with high mortality rates following from, for example, predation and cannibalism (Peterson and Kimmerer, 1994), food concentration (Koski et al., 2010) or temperature affecting the timing of the spring bloom (Dutz et al., 2010).

FUNDING

M.K. was partly funded through a European Community Marie Curie postdoctoral fellowship and Carlsberg Foundation, E.B. was funded through a European Community Marie Curie postdoctoral fellowship (MCFI-2002-01148) and the HARVEST project (no. 178477) of the research Council of Norway. The study was additionally financed by LIFECO EU Q5RS-2000-30183 (2001), CONWOY No. 2052-01-0034 (2002 and 2003) and No. 21-03-0487 to S.J. (2005) from the Danish Research Agency.

ACKNOWLEDGEMENTS

We wish to thank the captain and the crew on R/V DANA for the cruise operations, Kristine Arendt, Frank Hansen, Louisa Poulsen and Sanna Gärtner for various parts of the cruise data in 2001 or 2002, Nguyen Huu Trung for the help with sampling in 2002, and Andy Visser for cruise organization in 2002, 2003 and 2005.

REFERENCES

Andersen
M. C.
Nielsen
T. G.
Hatching rate of the egg-carrying estuarine copepod Eurytemora affinis
Mar. Ecol. Prog. Ser.
 , 
1997
, vol. 
160
 (pg. 
283
-
289
)
Anderson
T. R.
Pond
D. W.
Stoichiometric theory extended to micronutrients: comparison of the roles of essential fatty acids, carbon, and nitrogen in the nutrition of marine copepods
Limnol. Oceanogr.
 , 
2000
, vol. 
45
 (pg. 
1162
-
1167
)
Berggreen
U.
Hansen
B.
Kiørboe
T.
Food size spectra, ingestion and growth of the copepod Acartia tonsa during development: implications for determination of copepod production
Mar. Biol.
 , 
1988
, vol. 
99
 (pg. 
341
-
352
)
Bunker
A. J.
Hirst
A. G.
Fecundity of marine planktonic copepods: global rates and patterns in relation to chlorophyll a, temperature and body weight
Mar. Ecol. Prog. Ser.
 , 
2004
, vol. 
279
 (pg. 
161
-
181
)
Calbet
A.
Agustí
S.
Latitudinal changes of copepod egg production rates in Atlantic waters: temperature and food availability as the main driving factors
Mar. Ecol. Prog. Ser.
 , 
1999
, vol. 
181
 (pg. 
155
-
162
)
Cohen
R. E.
Lough
R. G.
Length–weight relationships for several copepods dominant in the Georges Bank-Gulf of Maine area
J. Northw. Atl. Fish. Sci.
 , 
1981
, vol. 
2
 (pg. 
47
-
52
)
Daan
N.
Bromley
P. J.
Hislop
I. G. R.
, et al.  . 
Ecology of North Sea fish
Neth. J. Sea Res.
 , 
1990
, vol. 
26
 (pg. 
343
-
386
)
Dutz
J.
Mohrholz
V.
van Beusekom
J. E. E.
Life cycle and spring phenology of Temora longocornis in the Baltic Sea
Mar. Ecol. Prog. Ser.
 , 
2010
, vol. 
406
 (pg. 
223
-
238
)
Edler
L.
Recommendations on methods for marine biological studies in the Baltic Sea—phytoplankton and chlorophyll
Baltic Mar. Biol.
 , 
1979
, vol. 
5
 (pg. 
5
-
37
)
Fossheim
M.
Primicerio
R.
Habitat choice by marine zooplankton in a high-latitude ecosystem
Mar. Ecol. Prog. Ser.
 , 
2008
, vol. 
364
 (pg. 
47
-
56
)
Fransz
H. G.
Colebrook
J. M.
Gamble
J. C.
, et al.  . 
The zooplankton of the North Sea
Neth. J. Sea Res.
 , 
1991
, vol. 
28
 (pg. 
1
-
52
)
Fransz
H. G.
Gonzalez
S. R.
Steeneken
S. F.
Metazoan plankton and the structure of the plankton community in the stratified North Sea
Mar. Ecol. Prog. Ser.
 , 
1998
, vol. 
175
 (pg. 
191
-
200
)
Haas
L. W.
Improved epifluorescence microscopy for observing planktonic micro-organisms
Ann. Inst. Océanogr.
 , 
1982
, vol. 
58
 (pg. 
261
-
266
)
Halsband
C.
Hirche
H. J.
Reproductive cycles of dominant calanoid copepods in the North Sea
Mar. Ecol. Prog. Ser.
 , 
2001
, vol. 
209
 (pg. 
219
-
229
)
Hamilton
S.
Hamilton
R. J.
Sewell
P. A.
Hamilton
R. J.
Hamilton
S.
Extraction of lipids and derivative formation
Lipid Analysis—A Practical Approach.
 , 
1993
Oxford University Press
(pg. 
13
-
63
)
Harris
R.
Interactions between diel vertical migratory behaviour of marine zooplankton and the subsurface chlorophyll maximum
Bull. Mar. Sci.
 , 
1988
, vol. 
43
 (pg. 
663
-
674
)
Hasle
G. R.
Sournia
A.
The inverted-microscope method
Phytoplankton Manual
 , 
1979
Paris
UNESCO
(pg. 
88
-
96
)
Hays
G. C.
Proctor
C. A.
John
A. W. G.
, et al.  . 
Interspecific differences in the diel vertical migration of marine copepods: the implications of size, color, and morphology
Limnol. Oceanogr.
 , 
1994
, vol. 
39
 (pg. 
1621
-
1629
)
Hays
G. C.
Warner
A. J.
Lefevre
D.
Long-term changes in the diel vertical migration behaviour of zooplankton
Mar. Ecol. Prog. Ser.
 , 
1996
, vol. 
141
 (pg. 
149
-
159
)
Hirst
A. G.
Sheader
M.
Are in situ weight-specific growth rates body-size independent in marine planktonic copepods? A re-analysis of the global syntheses and a new empirical model
Mar. Ecol. Prog. Ser.
 , 
1997
, vol. 
154
 (pg. 
155
-
165
)
Huntley
M.
Lopez
M.
Temperature dependent production of marine copepods: a global synthesis
Am. Nat.
 , 
1992
, vol. 
140
 (pg. 
201
-
242
)
Jansen
S.
Wexels Riser
C.
Wassmann
P.
, et al.  . 
Copepod feeding behaviour and egg production during a dinoflagellate bloom in the North Sea
Harmful Algae
 , 
2006
, vol. 
5
 (pg. 
102
-
112
)
Jónasdóttir
S. H.
Fields
D.
Pantoja
S.
Copepod egg production in Long Island Sound, USA, as a function of the chemical composition of seston
Mar. Ecol. Prog. Ser.
 , 
1995
, vol. 
119
 (pg. 
87
-
98
)
Jónasdóttir
S. H.
Visser
A. W.
Jespersen
C.
Assessing the role of food quality in the production and hatching of Temora longicornis eggs
Mar. Ecol. Prog. Ser.
 , 
2009
, vol. 
382
 (pg. 
139
-
150
)
Jónasdóttir
S. H.
Koski
M.
Biological processes in the North Sea: Comparison of Calanus helgolandicus and Calanus finmarchicus vertical distribution and production
J. Plankton Res.
 , 
2010
 
(in press)
Kiørboe
T.
Sabatini
M.
Scaling of fecundity, growth and development in marine planktonic copepods
Mar. Ecol. Prog. Ser.
 , 
1995
, vol. 
120
 (pg. 
285
-
298
)
Klein Breteler
W. C. M.
Gonzalez
S. R.
Influence of cultivation and food concentration on body length of calanoid copepods
Mar. Biol.
 , 
1982
, vol. 
71
 (pg. 
157
-
161
)
Klein Breteler
W. C. M.
Schogt
N.
Development of Acartia clausi (Copepoda, Calanoida) cultured at different conditions of temperature and food
Hydrobiology
 , 
1994
, vol. 
292/293
 (pg. 
469
-
479
)
Koski
M.
Kuosa
H.
The effect of temperature, food concentration and female size on the egg production of the planktonic copepod Acartia bifilosa (Short communication)
J. Plankton Res.
 , 
1999
, vol. 
21
 (pg. 
1779
-
1789
)
Koski
M.
Friis Møller
E.
Maar
M.
, et al.  . 
Degradation of appendicularian houses by pelagic harpacticid Microsetella norvegica and other agents
J. Plankton Res.
 , 
2007
, vol. 
29
 (pg. 
641
-
654
)
Koski
M.
Klein Breteler
W.
Dutz
J.
Seasonal changes in food quantity and quality of the common North Sea copepods Temora longicornis and Pseudocalanus elongatus: a bioassay approach
Mar. Ecol. Prog. Ser.
 , 
2010
, vol. 
399
 (pg. 
141
-
155
)
[PubMed]
Maps
F.
Runge
J. A.
Zakardjian
B.
, et al.  . 
Egg production and hatching success of Temora longicornis (Copepoda, Calanoida) in the southern Gulf of St. Lawrence
Mar. Ecol. Prog. Ser.
 , 
2005
, vol. 
285
 (pg. 
117
-
128
)
Mullin
M. M.
Production of zooplankton in the ocean: the present status and problems
Oceanogr. Mar. Biol. Ann. Rev.
 , 
1969
, vol. 
7
 (pg. 
293
-
314
)
Mullin
M. M.
Sloan
P. R.
Eppley
R. W.
Relationship between carbon content, cell volume and area in phytoplankton
Limnol. Oceanogr.
 , 
1966
, vol. 
11
 (pg. 
307
-
311
)
Müller-Navarra
D.
Lampert
W.
Seasonal patterns of food limitation in Daphnia galeata: separating food quantity and food quality effects
J. Plankton Res.
 , 
1996
, vol. 
18
 (pg. 
1137
-
1157
)
Nielsen
T. G.
Munk
P.
Zooplankton diversity and the predatory impact by larval and small juvenile fish at the Fisher Banks in the North Sea
J. Plankton Res.
 , 
1998
, vol. 
20
 (pg. 
2313
-
2332
)
Nielsen
T. G.
Løkkegaard
B.
Richardson
K.
, et al.  . 
Structure of plankton communities in the Dogger Bank area (North Sea) during a stratified situation
Mar. Ecol. Prog. Ser.
 , 
1993
, vol. 
95
 (pg. 
115
-
131
)
Ohman
M. D.
Wood
S. N.
Mortality estimation for planktonic copepods: Pseudocalanus newmani in a temperate fjord
Limnol. Oceanogr.
 , 
1996
, vol. 
41
 (pg. 
126
-
135
)
Peters
J.
Dutz
J.
Hagen
W.
Role of essential fatty acids on the reproductive success of the copepod Temora longicornis in the North Sea
Mar. Ecol. Prog. Ser.
 , 
2007
, vol. 
341
 (pg. 
153
-
163
)
Peterson
W. T.
Kimmerer
W. J.
Processes controlling recruitment of the marine calanoid copepod Temora longicornis in Long Island Sound: Egg production, egg mortality, and cohort survival rates
Limnol. Oceanogr.
 , 
1994
, vol. 
39
 (pg. 
1594
-
1605
)
Poulsen
L. K.
Kiørboe
T.
Vertical flux and degradation rates of copepod fecal pellets in a zooplankton community dominated by small copepods
Mar. Ecol. Prog. Ser.
 , 
2006
, vol. 
323
 (pg. 
195
-
204
)
Richardson
K.
Nielsen
T. G.
Pedersen
F. B.
, et al.  . 
Spatial heterogeneity in the structure of the planktonic food web in the North Sea
Mar. Ecol. Prog. Ser.
 , 
1998
, vol. 
168
 (pg. 
197
-
211
)
Richardson
K.
Visser
A. W.
Pedersen
F. B.
Subsurface phytoplankton blooms fuel pelagic production in the North Sea
J. Plankton Res.
 , 
2000
, vol. 
22
 (pg. 
1663
-
1671
)
Sabatini
M.
Kiørboe
T.
Egg production, growth and development of the cyclopoid copepod Oithona similis
J. Plankton Res.
 , 
1994
, vol. 
16
 (pg. 
1329
-
1351
)
Sargent
J.
McEvoy
L.
Estevez
A.
, et al.  . 
Lipid nutrition of marine fish during early development: current status and future directions
Aquaculture
 , 
1999
, vol. 
179
 (pg. 
217
-
229
)
Tönnesson
K.
Tiselius
P.
Feeding of the chaetognaths Sagitta setosa and S. elegans in relation to prey abundance and vertical distribution
Mar. Ecol. Prog. Ser.
 , 
2005
, vol. 
289
 (pg. 
177
-
190
)
Uttermöhl
H.
Zur vervollkommung der quantitativen phytoplankton-methodik
Int. Ver. Theor. Angew. Limnol.
 , 
1958
, vol. 
9
 (pg. 
1
-
38
)
Wesche
A.
Wiltshire
K. H.
Hirche
H. J.
Overwintering strategies of dominant calanoid copepods in the German Bight, southern North Sea
Mar. Biol.
 , 
2007
, vol. 
151
 (pg. 
1309
-
1320
)

Author notes

Corresponding editor: Roger Harris