1 Introduction

Euphausiids are the dominant taxa among zooplankters that inhabit coastal and offshore regions (Marshall 1979). They play an important role in the transport of organic matter and energy from the surface layer, where primary production takes place, to a deeper layer, and from zooplankton of the upper trophic level to nekton and even seabirds (Mauchline and Fisher 1969). They constitute 8–15% of the planktonic organism biomass that inhabits the epipelagic and mesopelagic waters of the Northwest Pacific (Aizawa 1974). In addition, they have species-specific distribution, water temperature preference and tolerance, day and night vertical movement patterns, and feeding habits (Lalli and Parsons 1997; Sogawa et al. 2013). These specificities have been studied in various topographical areas. For example, Euphausia pacifica (E. pacifica), which is found in most areas in the North Pacific (Feinberg et al. 2013), has different vertical distributions and physiological and life patterns in different regions from the subtropical to subarctic (Gómez-Gutiérrez et al. 2006; Iguchi and Ikeda 1995). Since euphausiids inhabiting different regions have specific correlations with ambient ocean conditions, identifying their ecological characteristics makes it possible to more clearly elucidate the relationship between the specific water mass and community structure (Sogawa et al. 2013). They are important secondary consumers in coastal ecosystems, major feeders of primary producer phytoplankton, and major food sources for many commercial fish species (Chae et al. 2008; Sugisaki and Kurita 2004).

The Yellow Sea is a semi-enclosed marginal sea of the Western North Pacific, located on the continental shelf (Li et al. 2016) between China and Korea. It has an average depth of 44 m, and its central region extends in a north–south direction parallel to land with a depth of ≥ 70 m (Wei et al. 2016). Owing to the particular topography, different water temperatures, and hydrodynamic factors, the seasonal thermocline in the Yellow Sea covers a cold water mass low in the water column, resulting in the formation of a unique cold water mass known as the Yellow Sea Cold Water Mass (YSCWM; Ho et al. 1959; Yu et al. 2006, Fig. 1). This water column is very firmly seated in the lower layer during summer, when the surface water temperature rises sharply (Zhang et al. 2008). The YSCWM begins to form in spring concurrent with thermocline development; it is most strongly developed from July to August before gradually weakening as the thermocline dissipates. In November, the thermocline and the YSCWM simultaneously disappear (Xu et al. 2016). The YSCWM has a considerable influence on the growth and propagation of phytoplankton that inhabit the Yellow Sea (Ho et al. 1959; Wang 2001; Yu et al. 2006; Zhang et al. 1996).

Fig. 1
figure 1

(redrawn from Wei et al. 2016)

Sampling stations in the Yellow Sea. The dashed line represents the general range of the Yellow Sea Cold Water Mass in summer

Full size image

The spatial distribution of E. pacifica in the Yellow Sea appears to vary seasonally, which is closely related to the developmental stage, seawater temperature, and concentration of chlorophyll a (chl-a; Sun et al. 2011; Yoon et al. 2000, 2006). In particular, adults inhabit water with temperatures below 10 °C in summer. This characteristic is similar to that of Calanus sinicus (Wang et al. 2003). The biomass and production of zooplankton larger than 1 mm when the YSCWM appeared were higher inside than outside the cold water mass, whereas zooplankton smaller than 1 mm showed opposite results (Huo et al. 2012). In previous studies, relatively large zooplankton, such as E. pacifica, were found in the YSCWM in summer because they avoid high temperature in the surface layer. The feeding characteristics of the zooplankton in the cold water mass has, however, not yet been studied.

Stable isotope analysis is useful for studying the nutritional pathways in a food web that understands organic matter transfer and energy flow from basal organisms to top predators (Fry 2006; Layman et al. 2012). One of the greatest advantages of stable isotope analysis in food web studies is that information on stable isotope ratios in organisms can provide time-integrated information rather than snapshots obtained from gut content analysis. The carbon-stable isotope ratio is suitable for identifying the primary carbon source underlying the food web (Peterson 1999), and the nitrogen-stable isotope ratio is a good tracer to determine the trophic position of organisms (Cabana and Rasmunssen 1996); it is most commonly used to understand the energy flow of food webs (Grey 2006). Furthermore, the isotopic mixing model, which utilizes a combination of stable isotope ratios for feeder and potential food sources, can be used to calculate the relative contribution of each food source for the consumer (Parnell et al. 2010).

In the present study, we investigate the feeding characteristics of E. pacifica at the YSCWM development stage by using carbon- and nitrogen-stable isotopes. The ecological importance of the YSCWM as the over-summering site of E. pacifica is discussed on the basis of the results.

2 Materials and methods

2.1 Collection of zooplankton including E. pacifica

Plankton sampling was conducted on the research vessel Eardo (Korea Institute of Ocean Science and Technology). The collection of E. pacifica was performed at four stations in the Yellow Sea in spring (April) prior to YSCWM formation and summer (August) from 2012 to 2014, when the YSCWM was strongly developed (Fig. 1). After confirming the distribution depth of E. pacifica by using a scientific echo sounder (DT-X, BioSonics Inc., USA), a conical net (1 m in diameter with mesh size of 330 μm) was set to the corresponding water depth with a blinker, set for more than 5 min and was then hauled vertically. In 2014, the collection of spring samples was not possible owing to a ship accident near the survey station, and sampling was conducted only in summer. Zooplankton, a potential food source of E. pacifica, were collected by vertical tow from the bottom to the surface with a zooplankton net (60 cm in diameter with mesh size of 200 μm) at the same stations for E. pacifica sampling in August 2014. The zooplankton were classified by size using pore sieves of 5, 2, 1, 0.5, and 0.2 mm. The length of E. pacifica was measured using a ruler in mm after collection. All samples were stored at −20 °C in a freezer in a pre-combusted glass jar until analysis.

2.2 Water environment and collection of size-fractionated plankton (particulate organic matter)

The physico-chemical properties of water such as temperature, salinity, fluorescence, and dissolved oxygen in summer (August) from four stations in the Yellow Sea were measured by using a conductivity, temperature, and depth (CTD) device (SBE 911plus, SeaBrid, USA). The subsurface chlorophyll maximum (SCM) layer was determined by using a fluorescence sensor attached to a CTD on a Rosette sampler. Water samples (40 L) were collected from the SCM layer and the layer 5 m above the bottom. The collected seawater was filtered by a 200-µm sieve to remove large-sized particles and was then passed through a pre-combusted Whatman glass fiber filter (GF/D) with a pore size of 2.7 µm and diameter of 25 mm to collect nanoplankton (2–20 µm) and microplankton (20–200 µm). The filter was also used to collect particles 2.7–200 µm in size. Seawater passing through the GF/D was filtered by using a Whatman GF/F with a pore size of 0.7 µm to collect particles 0.7–2.7 µm, which is similar to the size of picoplankton (0.2–2.0 µm). The filter samples were divided for carbon- and nitrogen-stable isotope analyses, and each sample was replicated three times. After filtration, the filter was placed in a Petri dish, blocked with a foil, and stored in an on-board freezer at −20 °C.

2.3 Stable carbon and nitrogen isotope analysis

All samples collected on board were transferred to the laboratory in an icebox containing dry ice and were stored in a deep freezer at −80 °C until pre-treatment for stable isotope analysis. Frozen samples were dried in a freeze dryer, and zooplankton samples were pooled by size, powdered, and homogenized using a mortar and pestle. E. pacifica samples from 2012 and 2013 were pooled and used for stable isotope analysis. In 2014, samples were individually analyzed. Homogenized biological samples were subsampled for carbon- and nitrogen-stable isotope analyses. Subsamples for carbon were used to remove inorganic carbon and lipids to prevent the interference of carbon other than the assimilated carbon in the whole body. Biota subsamples for carbon were soaked in 1 N HCl, shaken overnight for remove inorganic carbon, washed with ultrapure water three times to remove HCl, and dried. After the addition of chloroform/methanol (2:1, vol:vol), removed lipids were extracted by ultrasonication. The supernatant was discarded three times to completely remove the lipids and the residue was then dried. The filter paper samples were placed in a desiccator with 35% HCl and were vacuumed for one day to remove the inorganic carbon. They were then neutralized by adding NaOH and were kept in a desiccator in vacuum. The subsamples for nitrogen were used without extra pretreatment after the lyophilization of biota and filter samples. Each prepared sample was put in a tin capsule and analyzed for carbon- and nitrogen-stable isotopes by using a stable isotope mass spectrometer (Isoprime100, Isoprime Ltd., UK) connected to an element analyzer (EA3000, Eurovector, Italy) in a tin capsule. The analyzed values are expressed in terms of delta notation and permil (‰), and the stable isotope ratio defines the difference between the reference material and analytical sample, according to the following equation:

$$ \begin{aligned} \delta^{ 1 3} {\text{C }}\left( {{\text{or }}\delta^{ 1 5} {\text{N}}} \right) \permille & = \left[ {\left( {R_{\text{sample}} - R_{\text{standard}} } \right) - { 1}} \right] \times { 1}000 \\ R & =^{ 1 3} {\text{C}}/^{ 1 2} {\text{C }}\left( {{\text{or}}^{ 1 5} {\text{N}}/^{ 1 4} {\text{N}}} \right). \\ \end{aligned} $$

The reference material used Vienna Pee Dee Belemnite for carbon isotope analysis and atmospheric nitrogen for nitrogen isotope analysis. The substances CH-6 and N-1, certified by the International Atomic Energy Agency, were analyzed once every 12 samples for calibrating the sample values. Their standard deviation was within 0.2‰ for carbon and 0.3‰ for nitrogen.

2.4 Statistical analysis

The quantitative contribution of the expected food source of E. pacifica was calculated by using the Bayesian mixing model SIAR v. 4.2 (Stable Isotope Analysis in R; Parnell et al. 2010). The potential food sources entered into SIAR were size-fractionated plankton and zooplankton from different water layers at the same stations. Some size-fractioned zooplankton samples were analyzed by pooling because of their small amount. In such cases, the standard deviation was not able to be determined, and SIAR was calculated by inputting the standard deviation as zero. The overall stable isotope fractionation factors of the diet tissue generally average 0.4‰ for carbon and 3.4‰ for nitrogen (Post 2002). In the present study, the contribution of food source was calculated by applying 0.4 ± 0.17‰ for carbon and 2.3 ± 0.28‰ for nitrogen, as reported in an aquatic food web (McCutchan et al. 2003). The statistical significance of the isotopic values for each sample was assessed using the Student’s t test and one-way analysis of variance (ANOVA) in SPSS V12.0 software.

3 Results

3.1 Hydrographic physico-chemical profile data

In summer (August), the average surface water temperature at stations A03, A07, and D03 was 24 ± 0.9 °C and the water depth was 87, 72, and 92 m, respectively. However, that at D07, which had the shallowest water depth of 39 m, showed a relatively lower surface temperature (Fig. 2). The average temperature in the bottom layer of 4 stations was 12.4 ± 4.8 °C. In particular, stations A03 and D03, at depths > 80 m, showed a well-developed cold water mass where the temperature was 10 °C lower than that at the surface. The average salinity was 31.9 ± 0.2 at the surface layer and 32.8 ± 0.2 at the bottom layer. Stratification occurred as a result of the density difference between the surface and bottom layers, which is consistent with the depth of the thermocline. The maximum fluorescence occurred at an average depth of 26.3 ± 9.9 m, which is consistent with the bottom of the thermocline. The dissolved oxygen showed a similar vertical distribution pattern to that of fluorescence except for that measured at D07.

Fig. 2
figure 2

Vertical distribution of temperature, salinity, and fluorescence concentration at sampling stations a A03, b A07, c D03, and d D07 in August 2014

Full size image

3.2 Seasonal change in stable isotope ratios of E. pacifica

The nitrogen isotopes of E. pacifica were always approximately 2‰ higher in August than that in April in both 2012 and 2013 (Fig. 3). The carbon isotope value in April differed significantly between years (P < 0.05). However, no significant difference was found in August, and the mean value ranged from −21 to −20‰. The nitrogen isotope values in August in all three years were not significantly different each other (P > 0.05).

Fig. 3
figure 3

Biplot of δ13C and δ15N values (mean ± SD) of E. pacifica from 2012 to 2014 in the Yellow Sea (white: spring; black: summer; circle: 2012; inverted triangle: 2013; square: 2014)

Full size image

3.3 Stable isotope signatures of size-fractionated plankton

In A03 and D03 affected by cold water mass, the picoplankton of the SCM layer showed the lightest carbon-stable isotope values and picoplankton of the bottom layer showed the heaviest nitrogen isotope values (Fig. 4). The nano + microplankton showed significantly different carbon and nitrogen isotopic values (P < 0.05), respectively, in the SCM and bottom layer. However, the difference was lower than that of the picoplankton, and the nitrogen values in the bottom layer were slightly heavier than those in the SCM layer. In coastal waters A07 and D07, the picoplankton of the bottom layer showed the lowest carbon and highest nitrogen-stable isotope values among the size-fractionated plankton, and the nano + microplankton of the SCM layer showed the highest carbon and lowest nitrogen isotopic values. Overall, the size-fractionated plankton in the area affected by the cold water mass showed lighter carbon isotopic values than those in the coastal area.

Fig. 4
figure 4

Biplots of δ13C and δ15N values (mean ± SD) of size-fractionated plankton and E. pacifica at sampling stations in August 2014 (filled diamond: particulate organic matter; open circle: zooplankton; inverted triangle: E. pacifica. BOT bottom layer, SCM surface chlorophyll maximum layer)

Full size image

3.4 Stable isotope signatures of large-sized zooplankton and E. pacifica

The carbon-stable isotopic values of zooplankton and E. pacifica were approximately 2‰ lighter in the cold water mass (A03 and D03) than in coastal waters (A07 and D07); a similar result was also shown in the large-sized plankton (Fig. 4). In the coastal waters, the carbon-stable isotopic values of E. pacifica in A07 and D07 were not significantly different (P > 0.05), although the nitrogen values were differed significantly (P < 0.05). In addition, the nitrogen-stable isotopic values of zooplankton and E. pacifica were similar in A07, although the range was relatively broad in A03. The carbon and nitrogen-stable isotopic values of E. pacifica in the cold water mass showed no significant difference (P > 0.05). On the other hand, E. pacifica in the cold water mass showed the lowest nitrogen-stable isotope values among the collected zooplankton. The body length of E. pacifica was 14.0 ± 3.4 mm (n = 38), and there was no significant difference (P > 0.05) between the stations.

3.5 Diet sources of E. pacifica

Among the 7 components of size-fractionated plankton collected as potential food sources of E. pacifica, the zooplankton of 2.0–5.0 mm had a mean contribution of 42%, and that of 0.5–1.0 mm accounted for 25% of the contribution at station A03, which had cold water mass below the thermocline in summer (Fig. 5). At station D03, zooplankton > 5.0 mm had the highest mean contribution of 22%, followed by zooplankton of 2.0–5.0 mm, at 17%. At coastal station A07, the nano + microplankton of the SCM layer had the highest contribution, at 15%. At station D07, zooplankton of 0.2–0.5 mm had the highest contribution, at 20%. The major food sources of E. pacifica in the Yellow Sea at stations A03 and D03, the cold water mass areas, were relatively large zooplankton of 2.0–5.0 mm and > 5.0 mm, respectively, whereas at the coastal area, they fed on all sizes of zooplankton and nano + microplankton. Picoplankton played a minor role as a diet source for E. pacifica at all stations.

Fig. 5
figure 5

SIAR boxplots showing the proportional contribution of potential E. pacifica prey at sampling stations in August 2014. The gray scale (from light to dark) indicates 95, 75, and 50% confidence intervals

Full size image

4 Discussion

In this study, the major food source of E. pacifica in the YSCWM in summer was large-sized zooplankton of 2.0–5.0 mm, which was revealed by SIAR using stable isotope values. However, E. pacifica was omnivorous rather than carnivorous in the coastal waters, where the influence of cold water is low. According to a study by Kang et al. (unpublished data), the most abundant zooplankton species in same area in 2012 was copepod Paracalanus parvus. In addition, copepods Calanus sinicus, Oithona atlantica, O. similis, Ditrichocorycaeus affinis (Corycaeus affinis), Acartia hongi, A. omorii, A. hudsonica, and Sagittidae Aidanosagitta crassa (Sagitta crassa) were dominant species. The sizes of the dominant species were P. parvus, 0.6–1.3 mm; C. sinicus, 2.07–3.6 mm; O. atlantica, 1.11–1.43 mm; O. similis, 0.6–1.2 mm; D. affinis, 0.62–0.87 mm; Acartia sp., 0.659–2.1 mm; and Sagittidae, 10–90 mm (WoRMS Editorial Board 2017). P. parvus showed high density at mainly 0–20-m depths (Kang et al. unpublished data). On the contrary, C. sinicus was predominant in terms of biomass and were distributed in a relatively wide range of water depth (Kang et al. unpublished data). Their approximately 2–3-mm size classifies them as 2–5-mm zooplankton, the stable isotopes of which were analyzed in this study. Therefore, the major food source of E. pacifica in cold water mass is likely C. sinicus.

The diet of Euphausiid can be divided into three types: phytoplankton, zooplankton, and organic detritus (Mauchline and Fisher 1969). Euphausia species including E. pacifica are considered to be omnivorous filter feeders (Ohman 1984; Suh and Choi 1998). They appeared to be a selective omnivore in feeding experiments conducted with diatoms and copepods as dietary sources, and small components such as bacteria in the food web could not be fed on directly or effectively (Ohman 1984). These species feed on marine snow formed directly by gelatinous zooplankton, such as coagulated small particles of phytoplankton and granular excrement or mucilage prey structures (Alldredge and Silver 1988; Dilling et al. 1998). The grazing rates of E. pacifica on marine snow do not depend on food quality (Dilling et al. 1998). In general, stable carbon isotopes in living organisms can reflect the origin of their food source in food web studies. The picoplankton sample, which had the smallest particle size in this study, showed a difference of more than 4‰ in the δ13C value from E. pacifica in the SCM and bottom layers. E. pacifica did not appear to directly feed on picoplankton, and it was presumed that its role as a food source was minimal.

In a study on the grazing and metabolism of E. pacifica collected in the Yellow Sea, a culture feeding experiment using a Coulter counter showed that E. pacifica fed mostly on microzooplankton containing ciliates in August and September (Tao et al. 2015). In contrast, the oxygen consumption rate peaked in April and was four times higher than that in September, which is related to high breeding and feeding rates. Moreover, the O-to-N ratio of E. pacifica was highest in April, coinciding with spawning and the highest prey abundance, resulting in active metabolism (Sun et al. 2011; Tao et al. 2015). During September and December, the O-to-N ratio of E. pacifica was 90% lower than April with low phytoplankton concentrations (Tao et al. 2015). It generally indicates a stressed condition, such as starvation (Mayzaud and Conover 1988) and consequence of reduced respiration rates (Tao et al. 2015). However, most energy sources from September to December were proteins derived from a carnivorous diet containing microzooplankton. In the YSCWM in summer and autumn, the energy consumption in E. pacifica remained low. The O-to-N ratio of E. pacifica collected in April, September, and December showed a negative correlation with surface water temperature, and there was no significant correlation with chl-a concentration (Tao et al. 2015). Therefore, water temperature is an important factor in E. pacifica metabolism. In the present study, we calculated the contribution of predicted food sources to E. pacifica by using carbon- and nitrogen-stable isotopes. Zooplankton of the 2.0–5.0-mm size class at station A03, where the cold bottom layer was the most developed, contributed an average of 42.3% to its diet. Therefore, E. pacifica mainly fed on large-sized zooplankton. In previous studies, the potential food sources of E. pacifica ranged from small organic matter to microplankton (Tao et al. 2015). However, in the present study, we considered larger plankton as a possible food source. On the basis of dual stable isotope analysis, we determined that the contributions of mesoplankton and microplankton are considerable.

The optimum temperature of E. pacifica in the marginal sea of the western North Pacific was 11.4 °C, and they cannot survive more than one day at temperatures above 20 °C (Iguchi and Ikeda 1995). Higher summer temperatures may alter the balance of parameters that have inter-dependence between molting rates and chl-a, and between molting increment and temperature in metabolic responses. It may inhibit the growth of E. pacifica (Pinchuk and Hopcroft 2007). From summer to autumn, the surface water temperature in the Yellow Sea is above 20 °C. E. pacifica inhabiting the YSCWM exhibited decline in grazing, metabolism, and reproduction rates (Tao et al. 2015). The YSCWM and the water layer below the thermocline may provide conditions necessary for E. pacifica growth. In the present study, an acoustic scattering layer of E. pacifica identified by an echo sounder existed at depths close to the seabed in the cold waters but at 10–20 m above the seabed in coastal waters. It is believed that the activity of E. pacifica is significantly reduced in the bottom cold water mass, where the thermocline is strongly developed and the water temperature is less than 10 °C. The stable isotope values of E. pacifica for the 3-year period showed that the average carbon isotope values in summer were less variable from −21 to −20‰ compared with spring. The relatively small range in carbon isotope of E. pacifica might be attributed to the limited habitat available in the cold water mass in summer, which provides very limited diet sources such as zooplankton. However, E. pacifica in spring appears to have access to a variety of food sources that promote active feeding and reproductive activities, which are supported by the comparatively wide range in carbon isotope values in spring. The habitat of E. pacifica is restricted to the cold water mass in summer, which provides very limited diet sources such as zooplankton. In addition, the heavier nitrogen isotope signatures in summer compared with those in spring suggest that E. pacifica feeds more on carnivorous diet sources. Therefore, considering the current dual-isotope results, E. pacifica may feed mainly on zooplankton larger than 2.0 mm such as copepods C. sinicus even though they might be physiologically less active in the YSCWM in summer.

The stomach content analysis revealed that the feeding behavior of E. pacifica at the northeastern coast of Japan varies according to the surrounding food conditions (Nakagawa et al. 2001; Taki et al. 2002). In addition, the prey of this species differs seasonally from herbivorous feeding of diatoms in the spring to more carnivorous feeding of mainly copepods and detritus in the summer, fall, and winter (Endo 1981; Taki et al. 2002). The analysis of stomach contents and stable isotopes for adult Euphausia vallentini in the Southern Ocean revealed differing results in the two methods (Gurney 2000). The stomach contents showed high phytoplankton and low metazoan compositions, whereas stable isotope analysis showed a high omnivorous trend. Nakagawa et al. (2001, 2002) highlighted the importance of feeding on heterotrophic prey such as copepods in terms of the carbon weight calculation of heterotrophic prey and pigment content of autotrophic prey among E. pacifica prey. Sogawa et al. (2017) reported that no seasonal variation was noted in the nitrogen-stable isotopic values of E. pacifica with a standard deviation of 0.2‰ in the northeastern coast of Japan. Therefore, even though phytoplankton with frustles such as diatoms remain intact in the stomach contents, the effect of the phytoplankton spring bloom of Oyashio water on the trophic level fluctuation of E. pacifica would be small. In this study, E. pacifica in the Yellow Sea, with a seasonally varying water environment, showed heavier nitrogen isotopic values than those measured in spring, and the carbon-stable isotopic values were narrowed to a particular range. Therefore, it is suggested that food source of E. pacifica was limited in the YSCWM where E. pacifica escaped from warm surface water.

On the contrary, a study conducted in the southern part of the Yellow Sea in 2009 revealed that the nitrogen-stable isotope ratio of nitrate in seawater was 6.7 ± 0.8‰ (n = 61) in February and 6.9 ± 2.6‰ (n = 23) in July (Umezawa et al. 2014). Although a slight change was noted in the mean value between the two periods, it increased significantly at some stations. In this study, an increase in the nitrogen isotopic values of E. pacifica may be attributed to changes in primary producers and nitrogen sources. Therefore, it is necessary to understand the seasonal changes in isotopic values between primary producers and krill for more accurate interpretation.

5 Conclusion

We studied the feeding characteristics of E. pacifica in the YSCWM by using carbon- and nitrogen-stable isotopes. The stable isotope signature of E. pacifica inhabiting the cold water mass was confined to a specific range, which means that the food sources of E. pacifica are limited. Previous studies have reported that the physiological activity of E. pacifica in the cold water mass is reduced in summer. However, stable isotope analysis in the present study revealed that their grazing activity on relatively large zooplankton is still present in summer. Therefore, the YSCWM, which is strongly developed in summer, is a very important habitat for the survival of E. pacifica and provides shelter for avoiding the high-temperature seawater surface as well as a zooplankton diet source.