Se-Jong Ju · Hyung-Ku Kang · Woong Seo Kim · H. Rodger Harvey
Received: 10 September 2008 / Accepted: 17 March 2009 / Published online: 8 April 2009
© Springer-Verlag 2009
Abstract To better understand the feeding and reproduc- tive ecology of euphausiids (krill) in diVerent ocean envi- ronments, lipid classes and individual lipid components of four diVerent species of euphausiids from Northeast PaciWc (temperate species) and Southern Ocean (Antarctic species) were analyzed in animals from multiple life stages and sea- sons. The dominant krill species in the Northeast PaciWc Euphausia paciWca and Thysanoessa spinifera, were compared to the two major Antarctic species, Euphausia superba and E. crystallorophias. Analysis comprised total lipid and lipid classes together with individual fatty acid and sterol composition in adults, juveniles, and larvae.
Antarctic krill had much higher lipid content than their tem- perate relatives (10–50 and 5–20% of dry mass for Antarctic and temperate species, respectively) with signiWcant seasonal
variations observed. Phospholipids were the dominant lipid class in both temperate krill species, while neutral storage lipids (wax esters and triacylglycerols for E. crystalloro- phias and E. superba, respectively) were the major lipid class in Antarctic krill and accounted for up to 40% of the total lipid content. Important fatty acids, speciWcally 16:0, 18:19, 20:53, and 22:63, were detected in all four krill species, with minor diVerences between species and sea- sons. Detailed lipid proWles suggest that krill alter their lipid composition with life stage and season. In particular, larval Antarctic krill appear to utilize alternate food resources (i.e., sea-ice associated organisms) during austral winter in contrast to juveniles and adults (i.e., seston and copepods). Lipid dynamics in krill among krill in both sys- tems appear closely linked to their life cycle and environ- mental conditions including food availability, and can provide a more complete comparative ecology of euphausi- ids in these environmentally distinct systems.
Introduction
Euphausiids are one of the most abundant and widely dis- tributed animals in the pelagic marine environment and have adapted to radically diVerent environmental condi- tions, both physical and biological (Mauchline 1980). Some species (e.g., Euphausia paciWca) show a wide latitudinal distribution, while others (e.g., Euphausia crystalloro- phias) are restricted to small geographical areas (Mauchline and Fisher 1969). Although a few species of euphausiids reside in coastal waters, most are oceanic inhabitants. The limits of distribution for a particular species appear to be related to hydrographic conditions, with seasonal changes known to occur at the boundaries of their distribution (Brinton Communicated by U. Sommer.
S.-J. Ju
Deep-sea and Marine Georesources Research Department, Korea Ocean Research and Development Institute, P.O. Box 29, Ansan, Seoul 425-600, Republic of Korea H.-K. Kang
Marine Living Resources Research Department, Korea Ocean Research and Development Institute, P.O. Box 29, Ansan, Seoul 425-600, Republic of Korea W. S. Kim
Yeosu Exposition Supporting Task Force Team, Korea Ocean Research and Development Institute, P.O. Box 29, Ansan, Seoul 425-600, Republic of Korea H. R. Harvey (&)
Chesapeake Biological Laboratory,
The University of Maryland Center for Environmental Science, P.O. Box 38, Solomons, MD 20688, USA
1460 Mar Biol (2009) 156:1459–1473 mass distribution (temperature and current), but are also
linked to the resultant food availability, predation risk, and perhaps physiological conditions. Although euphausiid spe- cies may be herbivorous or carnivorous, many are omnivo- rous at some point in their life history (Ohman 1984;
Huntley et al. 1994; Atkinson et al. 2002; Meyer et al.
2002; Nakagawa et al. 2002), and each species must develop a means to cope with the seasonality in food abun- dance during periods of very low food supply (e.g., winter condition at high latitudes) or seasonal changes in the nature of the food (temperate and subtropical area), both of which ultimately impact survivorship and reproduction.
To investigate the response of individual krill species to their dynamic environments, we determined the detailed lipid composition of multiple life stages among four species of euphausiids from two substantially diVerent oceanic regimes. E. paciWca and Thysanoessa spinifera are the most abundant and ecologically important krill species in the Northeast PaciWc Ocean (Brinton 1976; Feinberg and Peter- son 2003 and references therein). E. paciWca is oceanic and cosmopolitan over a large area of the Northern PaciWc and often dominant along and seaward of the shelf break. In contrast, T. spinifera is more neritic and is abundant in shelf waters and the only euphausiid common inshore of the 150 m isobath. These two temperate species also exhibit reproductive contrasts: T. spinifera has been generally con- sidered as a spring/summer spawner whereas E. paciWca is primarily spawns in summer and shows high spatial and inter-annual variation (Smiles and Pearcy 1971; Feinberg and Peterson 2003). Two important Antarctic krill species include Euphausia superba which has great signiWcance in Southern Ocean ecosystem (Nicol and de la Mare 1993) and E. crystallorophias as the other common neritic krill species. The spawning season of E. superba extends from mid-November to April with major spawning activities in austral summer of January and February (Quetin et al.
1994). E. crystallorophias spawns in October–November, earlier than E. superba and generally before the onset of spring algal bloom (Harrington and Thomas 1987). Both Antarctic species primarily feed on phytoplankton during the austral spring and summer, but have been reported to switch their dietary sources (detritus, ice algae, or cope- pods) and/or adopt non-feeding strategies (reduced metabo- lism, shrinkage in size, or use of energy reserves) during both non-bloom periods and the ice covered winter (e.g., Atkinson et al. 2002; Meyer et al. 2002; Ju and Harvey
major species from the Northeast PaciWc Ocean and com- pared with recent observations made by ourselves and oth- ers on the seasonal changes in lipids among important Antarctic species. The overall aim was to investigate lipid composition among multiple life stages of the four euphau- siid species as a reXection of their seasonal responses to their dynamic oceanic habitats.
Materials and methods
Study area and sample collection
This study includes results obtained as a component of both the US Northeast PaciWc and Southern GLOBEC programs, which operated as multidisciplinary oceanographic research programs focused on the Northern California Current system along coastal Oregon and in Marguerite Bay located along the western Antarctic Peninsula, respectively. With the exception of larval E. superba, all krill were collected using 1 m2 MOCNESS (Multi Opening/Closing Net Envi- ronmental Sensing System) equipped with nine 333m mesh nets. Antarctic E. superba larvae were hand collected by diver from under the ice at the shelf break (Ross et al.
2004). Krill were sorted by species and size (with total length (TL: mm) deWned as length from anterior margin of the eye to the tip of the telson) using a dissecting micro- scope and individually transferred to pre-combusted 8 ml amber vials. Post larval krill were sorted into two distinct size groups based on total length to distinguish adults and sub-adults. Sub-adults for each species were deWned as follows; E. paciWca: 10 mm < TL < 15 mm, T. spinifera:
10 mm < TL < 20 mm, E. superba: 15 mm < TL < 30 mm, and E. crystallorophias: 15 mm < TL < 25 mm. This matu- rity grouping corresponds well with previous observations (e.g., Brinton 1976; Siegel 1987; Tanasichuk 1998a, b). All samples were immediately frozen and stored at ¡70°C until detailed analysis.
Lipid extraction and analysis
Before lipid analysis, wet mass (as either individuals or composites) was measured by thawing and brieXy rinsing the specimens with Wltered water. A subset of krill (n= 61–
200) were dried at 60°C for 48 h for measurement of water content and conversion of wet mass (WM) of animals to
Mar Biol (2009) 156:1459–1473 1461 for E. Superba, DM = 0.265£WM (n= 147, r2= 0.97),
for E. crystallorophias, DM = 0.276£WM (n= 61, r2= 0.96).
Total lipids were extracted from freshly thawed animals three times with a mixture of CH2CL2:MeOH (1:1) in sol- vent-washed 25 mL screw-cap glass/TeXon lined-cap test tubes with probe sonication as previously described (Har- vey et al. 1987). The three extracts were pooled, the solvent removed by rotary evaporation, and total lipid redissolved in CH2CL2:MeOH (2:1). Sub-samples of the total lipid extracts were split for both lipid class analysis and mea- surement of individual lipid components.
Lipid class composition in animals was determined by thin-layer chromatography with Xame ionization detection (TLC-FID) using an Iatroscan MK-V Analyzer (e.g., Hazel 1985; Volkman et al. 1986; Ju et al. 1997). BrieXy, aliquots (1–2l) of total extracts were spotted onto replicate S-III Chromarods and developed in hexane:diethyl ether:formic acid (85:15:0.2) for separation of major lipid classes. Phos- pholipids (PL) remained at the origin and were quantiWed as a class, but not identiWed individually. Lipid classes were each identiWed and calibrated using both individual and mixtures of commercial standards [phosphatidylcholine (PC) for PL, cholesterol for cholesterol (CS), non-adeca- noic acid for free fatty acids (FFA), triolein for triacylglyc- erol (TAG), and palmitoleic acid stearyl ester for wax esters (WE); Sigma-Aldrich Co.] with calibrations conducted for each individual lipid class over a range of concentrations.
Peak areas of individual peaks were integrated (HP Chem- Station) and quantiWed using an appropriate mixture run in parallel. Total lipid content was determined by summation of all lipid classes quantiWed by TLC-FID. Overall preci- sion for total lipid and major classes was §10% or better.
The quantiWcation of individual lipid components in total lipid content relied on well-established methods for the structural determination and quantiWcation in individual lipid components (e.g., Mannino and Harvey 1999).
BrieXy, internal standards (5-cholestane for sterols and alcohols and non-adecanoic acid for fatty acids) were added to subsamples the total lipid extract. Remaining solvent was dried under N2 gas, and the sample was saponiWed using 0.5 N methanolic KOH with gentle heating at 70°C for 30 min. After cooling and addition of water, neutral lipid fractions (sterols and alcohols) were partitioned three times with a mixture of hexane:diethyl ether (9:1) and removed. The neutral fraction was dried under N2 gas and treated with 50L of bis(trimethylsilyl) triXuoroacetamide (BSTFA) with 25% pyridine at 50°C for 15 min to convert free alcohols to their corresponding trimethylsilyl (TMS) esters. Polar lipid fractions containing fatty acids were par- titioned similarly following acidiWcation to pH 2, dried
(FAMEs) by using BF3-MeOH at 70°C for 30 min. Proce- dural blanks were processed in parallel. Fatty acid methyl esters were quantiWed by capillary gas chromatography (HP-5890-II GC) equipped with Xame ionization detection.
Separations were performed with a DB-5MS fused silica column (60 m length£0.32 mm internal diameter£ 0.25m Wlm thickness; J & W ScientiWc) using hydrogen as the carrier gas. A further subsample of FAME fractions were transesteriWed into picolinyl ester derivatives with the mixture of potassium tert-butoxide and 3-hydroxymethyl- pyridine (1:2) for unequivocal determination of double bond position in unsaturated fatty acids (Destaillats and Angers 2002). Structural identiWcation of all individual compounds was conWrmed by capillary gas chromatogra- phy–mass spectrometry (GC–MS; Agilent 6870 GC with Agilent 5973 MSD). The GC–MS was operated at 70 eV with mass range acquisition of 10–700 amu. The column and temperature programs for the GC–MS are similar to that described above for GC, with helium as the carrier gas and an injector temperature of 250°C.
Statistical analysis
All data were tested before statistical analyses to satisfy assumptions for homogeneity of variances and normality of residuals. To determine the diVerences of total lipid content among krill species and life stage, the protracted Fisher’s least signiWcant diVerence (LSD) was used. Multi-regres- sion analysis was used to determine whether length and mass of krill was diVerent between season and species (Littell et al. 1999). Principal component analysis (PCA) was also performed on fatty acid data of all krill samples using covariance matrix with SAS (Version 8.02). All fatty acid data were normalized before natural log transformed to satisfy assumptions for homogeneity of variances and nor- mality of residuals. The Wrst two component scores for each krill sample were calculated from the Wrst two principal component coeYcients of each variable. Cluster analysis was applied to group the krill samples using the Wrst two component scores obtained from PCA.
Results
Total lipid and lipid class
Over the seasonal cycle, the total lipid content of temperate krill species E. paciWca and T. spinifera ranged from 5–20 to 7–27% of DM, respectively (Fig.1a, b). Adult E. paci- Wca lipid reserves were consistently lower than T. spinifera, and also showed only minor seasonal variations. T. spinif- era exhibited a range in concentrations, which were ele-
1462 Mar Biol (2009) 156:1459–1473
total lipid content with maturation was observed for either species. In contrast with temperate species, E. superba and E. crystallorophias displayed higher total lipid content, and more pronounced seasonal variations ranging from 10–36 to 19–46% of DM, respectively (Fig.1c, d). Total lipid content was substantially higher in adult Antarctic krill than seen for temperate species. Concurrently, Antarctic species also showed greater Xuctuations in their total lipid content, which appeared to be closely linked with reproduction and food availability while seasonal variation of lipid content was lower for the temperate species.
The contrasts seen in total lipid was extended to the lipid classes observed among the four species. PL were the dom- inant lipid classes in E. paciWca and T. spinifera (greater than 50 and 40% of total lipids, respectively) with lower amounts of neutral lipids (Table1). Substantial TAG levels
either Antarctic species. Although E. superba is not known to produce WE, very minor amounts were detected (1.4–
3.3%; Table1) and there is one report of small amounts of WE observed in E. paciWca (Saito et al. 2002). Given these low levels, however, it is more likely that the WE observed here are derived from WE-containing prey (e.g., copepods), which E. superba is known to consume (Ju and Harvey 2004). It was also interesting that in all four krill species the absolute quantities of PL, usually considered to be struc- tural lipids, were elevated with increasing total lipid con- tent (Fig.2).
Fatty acids
All species of krill contained a wide range of saturated (SFA), monounsaturated (MUFA), polyunsaturated Fig. 1 Seasonal variation of total lipid content (in percent of dry mass,
DM) among investigated krill species and life stages (a Euphausia paciWca, b Thysanoessa spinifera, c Euphausia superba, and d Eup- hausia cystallorophias). Filled symbols with standard deviations are
comparative measures of Antarctic krill in spring and summer months from Hagen et al. (2001) and Kattner and Hagen (1998). Gray shaded areas indicate the winter season
Furciliae Juvenile Adult (M+F) Adult (only F) Adult (only M)
Reproduction 0
10 20 30 40 50
Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May
Lipid content (% DM)
Spring bloom
Months
Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Months
A
Lipid content (% DM)
0 10 20 30
Jan Feb Mar Apr May Jun July Aug Sep Oct Nov Upwelling
Reproduction
Dec Months
Jan Feb Mar Apr May Jun July Aug Sep Oct Nov Dec Months
Upwelling 0
10 20 30
Reproduction
B
C
Spring bloom Reproduction 0
10 20 30 40 50
D
Mar Biol (2009) 156:1459–1473 1463
than T. spinifera, the only exception being adult E. paciWca collected during winter (January). For Antarctic species, the total concentration of fatty acids ranged from 114.5§36.9 mg g¡1 DM in larvae of E. superba collected in winter (July–August) to 234.0§28.9 mg g¡1 DM in adult of E. superba collected in fall (April–May). For all four species, the 16:0, 16:17, 18:19 and 7, and 20:53
observed between species and seasons. Several fatty acids did show unique distributions; the 14:0 acid was high (>10% of total fatty acids) only in sub-adult and adult E. superba, while the 18:19 was elevated (32.5% of total fatty acids) in adult E. crystallorophias. The 22:63 acid was only abundant in temperate krill species. A number of branched fatty acids and the major components of calanoid Table 1 The relative abundance (% of total lipid content: mean) of lipid classes present in krill
Total lipid contents without standard deviations were measured from single composite samples
n* the number of individual animals used for analysis, n the number of composite samples used for analysis, – not detected
Sampling time January February April June August
Lipid classes Larvae (n= 1)
Adults (n* = 2)
Adults (n* = 1)
Adults (n= 3)
Larvae (n= 2)
Juveniles (n= 3)
Adults (n= 8)
Larvae (n= 11)
Juveniles (n= 5)
Adults (n= 18) (A) Euphausia paciWca
PL 60.9 55.3 69.9 58.1 58.0 53.8 76.7 57.7 74.4 72.9
CS 12.2 6.8 6.7 5.7 18.7 15.3 7.6 10.0 7.7 6.0
FFA 14.0 5.8 2.9 4.0 12.1 14.4 4.4 14.1 7.1 9.7
TAG 12.9 30.3 18.2 30.3 10.2 12.7 10.9 13.4 7.9 10.1
WE – 1.8 2.3 1.9 1.0 3.8 0.4 4.8 2.9 1.3
Total lipid content (% of dry mass)
10.1 18.1 (1.4) 8.9 8.0 (1.2) 7.1 (2.4) 9.0 (1.6) 8.3 (1.8) 7.1 (2.4) 8.4 (2.2) 8.1 (1.9)
Sampling time April June August
Lipid classes Adults
(n= 3)
Juveniles (n= 3)
Adults (n= 6)
Larvae (n= 3)
Juveniles (n= 2)
Adults (n= 9) (B) Thysanoessa spinifera
PL 61.6 49.2 55.7 68.6 49.5 71.7
CS 9.4 7.5 5.9 4.8 4.3 5.3
FFA 16.1 8.5 4.8 6.7 4.8 8.3
TAG 10.4 26.2 26.8 18.7 36.2 11.7
WE 2.5 8.6 6.8 1.2 5.2 3.0
Total lipid content (% of dry mass)
10.4 (2.4) 17.6 (0.5) 15.1 (4.6) 14.6 (4.8) 18.0 (1.7) 8.5 (2.7)
Species E. superba E. crystallorophias
Sampling season Fall (April–May) Winter (July–August) Fall (April–May) Winter (July–August) Lipid classes Adults
(n* = 10)
Larvae (n= 2)
Juveniles (n* = 15)
Adults (n* = 26)
Adults (n* = 5)
Larvae (n= 2)
Juveniles (n= 1)
Adults (n* = 19) (C) Euphausia superba and E. crystallorophias
PL 48.5 44.8 50.4 55.4 43.3 46.6 38.7 42.6
CS 3.3 10.6 1.9 4.8 1.5 3.3 2.9 1.1
FFA 0.5 2.4 0.2 0.1 0.7 – 1.1 –
TAG 46.1 38.9 45.6 48.3 4.7 – 3.4 –
WE 1.6 3.3 1.9 1.4 49.8 50.1 53.9 56.3
Total lipid content (% of dry mass)
29.1 (4.2) 11.2 (4.4) 18.9 (3.8) 23.8 (6.8) 43.3 (3.1) 20.0 (6.2) 30.4 28.3 (6.3)
1464 Mar Biol (2009) 156:1459–1473
were also found in small quantities (<5% of total fatty acids) among all krill species.
To allow a more robust comparison of these reXectors of feeding and lipid metabolism, PCA was applied to the suite of fatty acids encountered in individual animals or composite samples. Scores for the Wrst two principal com- ponents which provide a measure of total variability of fatty acids are shown in Fig.3a. The Wrst two principal components described 70% of the total variability, with the Wrst principal component (PC 1) accounting for 50%
of the total variation and the second principal component (PC 2) accounting for 20%. An examination of loadings among individual fatty acids revealed that the Wrst compo- nent was mainly due to 16:17 and PUFAs. Positive dis- placements in the PC 1 resulted from two long-chain PUFA’s (the 20:53 and 22:63), which originate from diatoms and dinoXagellates, respectively. Negative dis- placements in the PC 1 were mainly due to 14:0 and 16:17. The Wrst principal component (PC1: Axis 1) iso- lated Antarctic species with high neutral lipid content from temperate species with high PL content. The positive (upward) displacements in PC 2 were due largely to SFA (14:0 and 16:0). Negative displacements in the PC 2 were driven by MUFA (mainly 18:19 and 16:15). The sec- ond principal component (PC2: Axis 2) separated krill
present in multiple life stages of two species from North- east PaciWc are closely linked.
Fatty alcohols and sterols
The fatty alcohol and sterol distributions of all krill species are shown in Table3. Trace amounts of fatty alcohols were also present in three of the krill species. In E. crystalloro- phias, however, fatty alcohols were present in high concen- trations and diverse distribution, reXecting WE as a major lipid component. The 14:0 and 16:0 alcohols comprised more than 70% of total fatty alcohols in E. crystalloro- phias, with high amounts of the 20:1 also seen in larvae.
The presence of the isoprenoid alcohol phytol (3, 7, 11, 15- tetramethylhexadec-2-en-ol), originating from the side chain of chlorophyll a, in all krill conWrms their consump- tion of algal-related materials (e.g., sedimenting detritus, phytoplankton). Average concentrations of sterols were generally higher in larvae (more than 5.0 mg g¡1 DM) compared to juveniles and adults for most krill species and lifestage with the exception of adult E. paciWca sampled in April. As the dominant sterol of all crustaceans, cholesterol (cholest-5-en-3-ol) was in abundance and accounted for at least 71% of total sterols in all krill species. Although pres- ent in low abundance, a diverse suite of sterols was found Fig. 2 The relative composition
of major lipid classes (as percent of dry mass, DM) and the increments of major lipid classes with increasing total lipid content (in percent of dry mass, DM) in investigated krill species (a Euphausia paciWca, b Thysanoessa spinifera, c Euphausia superba, and d Euphausia crystallorophias)
TAG= 0.52*TL -1.58 R2=0.91 PL = 0.47*TL + 0.65
R2= 0.91
0 5 10 15 20 25
0 10 20 30 40 50
Total lipid (% DM)
0 10 20 30 40 50
Total lipid (% DM)
Lipid class (% DM)
WE= 0.63*TL -2.74 R2=0.88
PL= 0.35*TL + 2.31 R2=0.85
0 5 10 15 20 25
A
TAG= 0.39*PL -1.93 R2=0.88 PL= 0.50*TL + 1.03
R2=0.77
0 2 4 6 8 10 12
0 5 10 15 20 25 0 5 10 15 20 25
Lipid class (% DM)
TAG (n=39) PL (n=39)
TAG = 0.48*TL -2.87 R2=0.84 PL = 0.39*TL + 2.09
R2=0.84
0 2 4 6 8 10 12
B
D C
TAG (n=23) PL (n=23)
TAG (n=56) PL (n=56)
WE (n=25) PL (n=25)