Introduction
Our knowledge of lipid biochemistry of marine Antarctic copepods is relatively recent compared to Arctic species and remains limited to oceanic areas. The earlier studies by Reinhard & Van Vleet (Reference Reinhard and van Vleet1986) and Hagen (Reference Hagen1988) surveyed the lipid class composition of the six major species (Calanus propinquus Brady, Calanoides acutus Giesbrecht, Rhincalanus gigas Brady, Metridia gerlachei Giesbrecht, Euchaeta antarctica Giesbrecht and Euchirella rostamagna Wolfenden) and the fatty acid structure of some of them. More recent work (see review by Lee et al. Reference Lee, Hagen and Kattner2006) has brought further information on the triacylglycerol (TAG) and wax ester (WE) structure in relation to seasons and life cycle variability.
However, to our knowledge, the small neritic species of both Antarctic and sub-Antarctic areas have received little attention despite their ecological significance (Tucker & Burton Reference Tucker and Burton1990, Ward & Shreeve Reference Ward and Shreeve1999). In the Iles Kerguelen, the zooplankton community is largely dominated by the small clausocalanid copepod species, Drepanopus pectinatus Brady, which displays very high spring and summer biomass and plays a key role in energy transfer towards higher trophic levels (Razouls & Razouls Reference Razouls and Razouls1988, Reference Razouls and Razouls1990, Razouls et al. Reference Razouls, Koubbi and Mayzaud1996). Throughout the year, a succession of four generations with different duration has been proposed by Razouls & Razouls (Reference Razouls and Razouls1990), but information on the energetic content and lipid composition of each generation is totally unknown. As indicated by Kattner et al. (Reference Kattner, Hagen, Lee, Campbell, Deibel, Falk-Petersen, Graeve, Hansen, Hirche, Jonasdottir, Madsen, Mayzaud, Muller-Navarra, Nichols, Paffenhoffer, Pond, Saito, Stubing and Virtue2007) in their perspectives paper “there is a general need for comprehensive field studies on lipid of zooplankton and the seasonal variability of these changes”.
During a general survey of the zooplankton community in Kerguelen, populations of Drepanopus pectinatus were sampled seasonally in order to examine lipid content, composition and structure in relation of both food interactions and life cycle stage.
Material and methods
Zooplankton samples were collected in the Bay of Morbihan by vertical hauls from bottom to the surface at central stations (see Fig. 1). Sampling was carried out from January 1996–February 1997, at monthly or bimonthly intervals depending on seasons for lipid characteristics. According to Razouls & Razouls (Reference Razouls and Razouls1990) the biological cycle of Drepanopus pectinatus comprised four generations: G1 October–December, G2 December–mid February, G3 end February–mid April, and G4 April–September.
Fig. 1 Map of sampling sites in the Bay of Morbihan of the Iles Kerguelen.
Consecutive temperature and water samples were taken using reversing temperature probes (Orca RTM T709) at two depths (0 and 30 m). Samples of 1 litre of surface water were filtered on GF/F filters for measurements of particulate chlorophyll a at three depths (0, 30, 80 m) using a Turner Design 10 fluorometer (Lorenzen & Downs Reference Lorenzen and Downs1966). Mesozooplankton were sampled with a 200 μm WP2 net by vertical hauls from bottom to surface. The zooplankton were immediately diluted in a plastic cooler with surface seawater and brought back to a laboratory cold room set to in situ temperature. For seasonal lipid studies, groups of 200–300 C6 females (with few C5 in autumn and winter) were sorted, immediately deep frozen and kept at -80°C under nitrogen and transported by air shipment to France on dry ice every 3–4 months. A subsample of female Drepanopus pectinatus was used for dry weight and total lipid determination. For dry weight, individuals (10–15) were placed on aluminium pans and dried at 60°C to constant weight. Weights were measured on a Cahn electrobalance.
On one date in January, D. pectinatus was sorted to stages to check the influence of ontogeny on the lipid composition. However, the high number of individuals needed for complete analyses (800–900 for younger stages, 200 for adults), made it difficult to obtain totally pure samples. Hence, a subsample of 50 individuals for each stage was fixed and counted under microscope: group C3 contains 10% of C4 (in number), group C4 contains 5% of C5, and group C5 contains 3% of females.
Lipid extraction and determination
Lipids were extracted upon arrival in France according to the method of Bligh & Dyer (Reference Bligh and Dyer1959). After solvent evaporation at high vacuum, the extracted lipids were weighed in tared vials on a precision balance (± 100 μg) to evaluate the content of total lipids. The extracts were then placed under nitrogen at -80°C until further analyses, usually within four months. The very low levels of free fatty acids were taken as evidence for proper lipid preservation.
Lipid classes were quantified after chromatographic separation coupled with FID detection on an Iatroscan Mark V TH 10 (Ackman Reference Ackman1981). Total lipid extracts were applied to Chromarods SIII using microcapillaries (1 μl) and analysed in duplicate. Neutral lipids were separated using a double development procedure with the following solvent systems: n-hexane: benzene: formic acid 80:20:1 (by volume) followed by n-hexane: diethylether: formic acid 97:3:1.5 (v/v). Calibration was achieved using either commercial standards (Sigma) or lipid extract (WE) from D. pectinatus separated by column chromatography (Mayzaud et al. Reference Mayzaud, Albessard and Cuzin-Roudy1998).
Fatty acid methyl esters of each lipid class were prepared with 7% boron trifluoride in methanol (Morrison & Smith Reference Morrison and Smith1964). Wax esters fatty alcohols were acetylated using acetic anhydride (Ackman et al. Reference Ackman, Hooper, Epstein and Kellchner1972). Gas liquid chromatography (GLC) of all esters was carried out on a 30 m length x 0.32 mm internal diameter quartz capillary column coated with Famewax (Restek) in a Perkin-Elmer XL Autolab gas chromatograph equipped with a flame ionization detector. The column was operated isothermally at 190°C for methyl esters and 200°C for alcohol acetates. Helium was used as carrier gas at 7 psig. Injector and detector were maintained at 250°C. Individual components were identified by comparing retention time data with those obtained from authentic and laboratory standards. In addition to the examination of esters as recovered, a part of all ester samples was completely hydrogenated and the products examined qualitatively and quantitatively by GLC. The level of accuracy is ± 5% for major components, 1–9% for intermediate components and up to ± 30% for minor components.
Data treatment
Non parametric Kolmogorov-Smirnov tests were used to evaluate significance between mean values using Statgraphics XV software.
To clarify the seasonal evolution of successive generations in relation to trophic interactions, the fatty acid profiles of the wax esters and triacylglycerols were subjected to a factorial correspondence analysis (Gower Reference Gower1987). The analysis was performed on a reduced data matrix transformed to relative frequencies and scaled so that each row (or column) can be viewed as a row (or column) of conditional probability distribution. The fatty acids were used to fill the matrix of variables and the different sampling dates were used as observations. Distances between fatty acid profiles were computed with a χ2 metrics. This distance gives symmetry to the two sets of data (active variables and observations) so that each factorial axis associated to the cluster of variables corresponds to a factorial axis of the cluster of observations. Thus, it was possible to represent simultaneously descriptors and observations on the plane defined by the factorial axes. Graphs of projections retained those variables which displayed more than 1% total contribution to axes 1 and 2. Hierarchical clustering on the fatty acids descriptors was carried out using reciprocal neighbour classification (Lebart et al. Reference Lebart, Morineau and Piron1995). Computation was made using the SPAD 5.5 software (DECISIA).
Results
Annual changes in environmental conditions
Temperature in the Kerguelen waters followed a very seasonal pattern with winter lows (July–August) between 2 and 3°C and maxima between 6 and 8°C from January–April (summer). Over the period considered the maximum recorded at both depths were higher in 1997 than in 1996 (Fig. 2 top). Chlorophyll also showed a strong seasonal signal but with a large interannual variability (Fig. 2 bottom) with a September spring bloom and succession of summer blooms in February and March 1996, and blooms of lower intensity in December, January and February 1997.
Fig. 2 Seasonal changes in sea surface and mid depth temperature and particulate chlorophyll at three depths at sampling sites.
Seasonal changes in size, dry weight and lipid content
No attempt was made to redefine the population dynamics and the scheme with four generations per year was used as framework of interpretation in the present study. Dry weight of C5-adult stages showed maximum values in spring and summer (Fig. 3) following the periods of phytoplankton abundance. Changes in dry weight were significantly related to the changes in total lipid content (F 1,27 = 12.18, P = 0.0017), which displayed similar seasonal pattern with summer high and winter low concentrations (Fig. 3). These changes are associated with changes in size, with maximum value in spring–summer and minimum in winter (Table I).
Fig. 3 Seasonal variations of adult female dry weight and lipid content per individual.
Table I Ranges in dry weight, size, lipid content and lipid class composition for the different generation (G) of female Drepanopus pectinatus.
nd = not determined, tr = trace.
Seasonal changes in lipid class composition.
The range of total lipid content and dry weight of females D. pectinatus for each generation are presented in Table I, together with the corresponding range in composition of lipid classes. WE dominated lipids at all times with mean percentages ranging from 65 to 73% in winter (G4) and summer and autumn (G2 and G3) respectively. TAG showed low contributions to the lipid pool with percentages ranging from 9–16% in winter (G4) and minimum values (0–10%) in summer (G2). Polar lipids (PL) remained the second constituent and ranged throughout the seasons from 14% (spring low G1) to 47% (winter high G4) of total lipids. Cholesterol and free fatty acids were always low at the limit of detection. The large range in the content of lipid within generation is indicative of the high degree of variability. The differences between means within generation were usually not significant (P > 0.05) suggesting that, in the present case, generation may not be the proper time scale to address the seasonal variation of lipid reserves.
Changes in total lipids were linearly correlated to both polar and reserve lipids (Fig. 4). WE dominated the lipid dynamics with a highly significant regression (WE = -1.96 + 0.75 Liptot; P = 0.0001, F 1,28 = 199.1). TAG were also significantly related to total lipids but with a lower slope (TG = -2.19 + 0.21 Liptot; P = 0.0004, F 1,14 = 22.99). Contrary to the expected trend, PL lipids were significantly related to total lipids with a slope similar to that of triacylglycerols (PL = 3.89 + 0.10 Liptot; P = 0.0001, F 1,28 = 20.63).
Fig. 4 Relation between total lipids and concentration of the main lipid classes for adult females.
Seasonal changes in the main three lipid classes as percent total lipids (Fig. 5) illustrated the inverse pattern between PL and WE with high WE percentages in spring and summer and decreasing levels through winter. TAG remained low without any clear seasonal pattern. Interestingly, there was no relation between the catabolism of neutral lipids and the peak of reproduction between November and March.
Fig. 5 Seasonal variability in lipid classes as percentage of total lipids for adult females.
The influence of ontogeny was monitored on one occasion in January using groups of dominant growth stages (Table II). Because of the limited amount of material, no attempt was made to weigh total lipid to avoid sample degradation. Early C3 and C4 stages showed little accumulation of wax esters and relatively high percentages of TAG and PL. Subadult C5 was the stage of WE accumulation with values exceeding 70% of the total lipids. Triacylglycerols fell below 10% and PL below 15%. Females retained high percentages of wax esters. Changes in dry weight showed maximum increase between stage C4 and C5.
Table II Drepanopus pectinatus. Dry weight (μg/ind) and lipid class composition (% total lipids per growth stage). WE = wax esters, TAG = triacylglycerols, FFA = free fatty acids, Chol = cholesterol, PL = phospholipids.
Fatty acid and fatty alcohol constituents of wax esters
The fatty acid composition of WE was dominated at all times by the monoenes (sum ranging from 47–75% of the total fatty acids) with oleic acid (18:1n-9) as the main constituent and to a lesser extent palmitoleic acid (16:n-7) (Table III). Saturated acids were dominated by myristic acid (14:0) with maximum percentages in early winter. Palmitic (16:0) and stearic (18:0) acids were generally low with values < 1% of total fatty acids. PUFA with 5 and 6 double bonds showed a strong seasonal signal with minimum values for the winter generation and maximum values for the summer period (Table III). Polyenes with four double bonds were the dominant polyunsaturated (PUFA) fraction with 18:4n-3 ranging from less than 3% in winter and interesting high levels of 10% during spring and summer. Substantial amount of 20:5n-3 (10–15%) was recorded during summer while smaller, but significant, amounts of 16:4n-1 was abundant in summer. Dienoic acids were essentially 16:2n-4 and 18:2n-6 while trienoic acids were dominated by 16:3n-4 and 18:3n-3. The changes in n-3 PUFA showed marked seasonal changes with a decreasing trend in autumn, minimal values in winter and increasing percentages throughout spring and summer. Changes in n-6 PUFA were less pronounced with lower percentages in January and February. The resulting ratio ranged from less than 0.1 to 0.5 with maximal values during the winter months.
Table III Drepanopus pectinatus. Fatty acid composition of the wax esters (% of total methyl esters).
The fatty alcohol fraction of the WE were dominated at all time by C14 and C16 alcohols (Table IV) which accounted for over 90% of the total alcohols. Minor constituents included 18:1n-5 alcohol which displayed maximal values in winter (August–September) and iso-, anteiso-branched alcohols, which showed a clear seasonal pattern with minimum values in summer and maxima in winter. Unsaturated alcohols with two and three double bonds were occasionally detected but always in percentages lower than 0.2% (not shown).
Table IV Drepanopus pectinatus. Fatty alcohol composition of wax esters (% of total alcohols).
Fatty acid composition of TAG
The fatty acid composition of the TAG fraction was dominated by saturated and monoenoic acids in similar proportion ranging from 25–42% and from 23–48% for saturates and monoenes respectively (Table V). Palmitic (16:0), myristic (14:0) and to a minor extent stearic (18:0) acids made the bulk of the saturated acids while oleic (18:1n-9) and palmitoleic (16:1n-7) acids dominated the monoenoic acids. Among PUFA 18:4n-3, EPA (20:5n-3) and to a minor extent DHA (22:6n-3) were the main constituents. Minor PUFA included 16:3n-4, 16:4n-3, 16:4n-1, 18:3n-3 which are considered as diet markers. Over time, 14:0 and 18:0 dominated in winter and early spring (August–early November) while 16:0 showed an opposite trend. PUFA (n-3) showed maximum percentage in December (late spring) and minimum values in winter (August–September). PUFA (n-6) displayed a similar pattern of changes but with lower intensity yielding a Σ PUFA n-6/Σ PUFA n-3 ratio which varied from 0.05 in summer to 0.25 in winter (Table V). C20 and C22 monoenes failed to show a clear seasonal pattern with higher percentages between May and July (autumn–early winter) and very low values for the rest of the survey period.
Table V Drepanopus pectinatus. Fatty acid composition of triacylglycerols (% total fatty acids).
Fatty acids of phospholipids (PL)
Phospholipids showed relatively weak seasonal changes with DHA and to a lesser extent EPA as the major fatty acids (Table VI). Saturated acids were the second largest contributor with palmitic acid as the major constituent (> 22% total fatty acids) associated with lower levels of myristic and stearic acids. Monoenoic acids were always intermediate with percentages < 11% and oleic acid (18:1n-9) as the main component. Other PUFA included 18:4n-3 and minor percentages of other n-3 acids (Table VI). The only seasonal pattern concerned DHA concentration which was minimal in winter and maximum in spring and EPA and palmitic acid which showed a reverse pattern of changes. Among the branched acids, iso15:0, iso17:0 and anteiso17:0 displayed a clear winter maximum.
Table VI Drepanopus pectinatus. Fatty acid composition of phospholipids (% total fatty acids).
Fatty acid composition of ontogenetic stages
Changes in fatty acid composition from copepodite C3 to adult stages are presented in Table VII for both total polar and neutral lipids. Polar lipids exhibited a major decrease in total saturates and monoenes with growth stage and conversely increasing percentages in PUFA. Copepodite C3 displayed the smaller level of DHA and EPA and maximum values for C5 and adult female. The same applied at a smaller scale for 18:4n-3. Conversely, palmitic and stearic acids, and to a minor extent palmitoleic and oleic acids, were maximum for C3 and decrease sharply as early as C4 stage. The TAG pattern displayed was different with increasing percentages of total saturates and dienoic acids, decreasing levels of total monoenoic acids while PUFA did not show a clear pattern of variation. Palmitic acid, palmitoleic and DHA displayed the largest increase in percentages, while oleic acid, stearic acid, linoleic acid decrease mainly between C3 and C4 stages. The same pattern was observed for 18:4n-3 but with a decrease between C4 and C5. EPA and DHA showed little changes with no clear pattern. WE showed increasing percentages of total monoenoic acids and dienoic acids but no clear pattern for the sum of saturated acids. Conversely, total PUFA showed decreasing percentages with increasing copepodite stages. The decrease was essentially associated with 18:4n-3, DHA and linoleic acid, which dropped either between C3 and C4 or between C4 and C5 and to a minor extent with oleic, palmitic and stearic acid. Increasing percentage concerned essentially 16:4n-1, 16:3n-4, 16:2n-4 and myristic acid.
Table VII Changes in fatty acid composition (% total fatty acids) of the different classes of lipids for copepodite stages and adult females of Drepanopus pectinatus. PL = phospholipids, TG = triglycerides, WE = wax esters.
To illustrate the probable changes in trophic interactions with growth stage, a factorial correspondence analysis (FCA) was performed on the fatty acid composition of the TAG from stage C3 to adults. Due to the limited number of cases the first two factorial axes accounted for 96% of the total variance and the link between growth stage and characteristic fatty acid grouping can be derived from a hierarchical clustering on factorial score from all three axes. The results presented in Fig. 6 showed a strong opposition on axis 1 (85% total inertia) between stage C3 and older stages with C3 associated with saturated and branched acids as well as (n-3) 16 PUFA and 18:4n-3. This suggests feeding on small-size particles (live and detritic). Later stages showed a clear association with diatom bloom type material with C16 PUFA of the (n-4) and (n-1) families, 16:1n-7 and EPA. Interestingly, the second axis (11% total inertia) discriminate C4 from C5 and adults suggesting a differential use of the diatom bloom material, but similar trophic relationships for C5 and adults.
Fig. 6 Hierarchical cluster analysis on the score from the factorial correspondence analysis on triglyceride fatty acids from the different copepodite stages (C3 to adults). The clustering illustrates the results of the FCA considering the projections on the first three axes and summarizes the associations between growth stages and fatty acids in all three directions of the factorial space.
Seasonal changes in fatty acid composition and trophic interactions of adult females
To clarify the importance of season, FCA was performed on the fatty acid constituents of the wax esters and triacylglycerols fractions. For WE, three factorial axes are needed to explain 75.3% of the total variance (axis 1 = 32.5%, axis 2 = 26.3%, axis 3 = 16.5%). Projection on the factorial plan defined by the first two axes opposes the summer period (January–February) to the spring one (November–December) on the first axis (Fig. 7 top). Spring and summer individuals are opposed to the winter ones (July–September) on the second axis. The third axis (not shown), singles out the 19/11 sampling date. The seasonal path derived from the wax esters showed a transition from summer to autumn and to winter along the second axis. An internal shift within wax esters fatty acids of the winter individuals can be seen along the first axis from July–September. In terms of fatty acid descriptors the opposition on the first component between summer and spring is characterized respectively by 16 PUFA, EPA versus a group of n-3 and n-6 PUFA (16:4n-3, 18:3n-3, 20:3n-3, 20:2n-6). The winter generation on the second axis is associated with saturates, monoenoic acids and 16:3n-3. The winter changes suggested that August–September wax esters showed increasing proportion of some PUFA (16:2n-4, 20:4n-3).
Fig. 7 Factorial correspondence analysis on the seasonal changes in fatty acid composition for wax esters (top) and triglycerides (bottom). Top: combine projections of wax ester fatty acids and sampling dates. Bottom: single projection of triglycerides sampling dates. The figure shows the ordination of the variables and observations in the plane of axis 1 and axis 2. The major trend of seasonal variation is illustrated by the dashed line arrows joining the sampling dates. Fatty acids illustrate for each season the key descriptors associated with the time periods.
The projection of the TAG fatty acid confirmed a globally similar seasonal pattern except that the factorial analysis suggested a different relationship between variance and processes involved. Indeed, axis 1 accounted for 41.6% of the total variance and singles out the shift in triglyceride composition between early and late winter (Fig. 7 bottom). The second axis opposes the spring–summer individuals to the winter one but with only 26.8% of the total inertia. The fatty acid descriptors are relatively similar to the one recorded for the wax esters with spring associated with (n-3) 16, 18 PUFA and EPA, and summer characterized by (n-1) (n-4) C16 PUFA and 14:1 (not shown). The triacylglycerols from winter individuals are divided into two groups: June–July associated with 22:1 and 20:1 acids and August–September associated with 14:0, 16:0, C16 and C18 monoenes, and branched fatty acids.
Discussion
The copepod Drepanopus pectinatus is considered endemic to Kerguelen and constitutes 70–98% of the total copepod population (Razouls & Razouls Reference Razouls and Razouls1988). Based on cohort analysis, its biological cycle has been shown to comprise four generations of different durations (44–49 days in summer to 151 days in winter) which vary in relation to temperature and food availability, with populations growing from G1 to G3 and slower development or possibly arrest of moulting for the winter generation G4 (Razouls & Razouls Reference Razouls and Razouls1990).
In copepods, lipid accumulation has been classically related to seasonality in food supply and reproductive metabolism (see review by Lee et al. Reference Lee, Hagen and Kattner2006). Large accumulations of wax esters in copepods or euphausiids are usually related to their overwintering strategy (with or without diapause), while the link with reproduction is more complex with either species fuelling their egg production with accumulated lipids or species using food - derived lipid to cover their reproductive needs. As observed by Lee et al. (Reference Lee, Hagen and Kattner2006) most data at present concerns species from polar or subpolar regions, i.e. large Calanus type copepods and little is known concerning the lipid dynamics of lipid-poor or lipid-rich small neritic copepods. An earlier report on different small species from the North Sea, such as Pseudocalanus elongatus, Acartia clausi and Centropages hamatus, demonstrated that lipid content was directly related to spring and summer blooms of phytoplankton (Kattner et al. Reference Kattner, Krause and Trahms1981). Most of these species have two or more generations per year and the relationship between lipid dynamic, generation succession, trophic environment and reproductive strategy is largely unknown, though it has generally been assumed that egg production was directly or indirectly related to intake or synthesis from available food source (Corkett & McLaren Reference Corkett and McLaren1969, Bautista et al. Reference Bautista, Harris, Rodriguez and Guerrero1994).
Our data on Drepanopus pectinatus illustrated that, although this is a multigenerational species, intra-generation variability of lipid class composition dominated. Indeed, spring and summer generations showed large fluctuations while the winter generation showed a strong seasonal trend throughout the period. As a result the seasonal lipid dynamics did not seem to be linked to any single generation but rather to the general pattern with a strong winter decline. A possible general explanation could be that the lipid dynamics of a given generation is driven by the accumulation of the previous generation of females. Hence, spring and summer variability in female lipid class composition could be linked to successive phases of neutral lipid accumulation and catabolism related to the high reproductive rate taking place during this period of time (Alonzo et al. Reference Alonzo, Mayzaud and Razouls2001). Late autumn–winter decrease in lipid reserves is probably a response to both low food conditions and lower level of reproduction entirely based on the internal pool of lipid reserves (Alonzo et al. Reference Alonzo, Mayzaud and Razouls2001). Although total lipid content is moderate (30–40% dry weight) when compared to large Calanus species (Lee et al. Reference Lee, Hagen and Kattner2006), the seasonal dynamics of lipids was related to the accumulation of wax esters in spring and summer (reproductive period). The coincidence between high phytoplankton biomass and high wax esters content does not mean that lipid accumulation controlled egg production since Alonzo et al. (Reference Alonzo, Mayzaud and Razouls2001) established that D. pectinatus used ingested phytoplankton to reduce the consumption of their internal lipid reserves. The seasonal changes in total lipid and WE suggested that for multigeneration species, the overwintering pattern with high lipid accumulation does not apply since winter individuals showed the smallest size and lowest lipid levels. However, the relatively high proportion of WE in late autumn and the decreasing trends as winter developed is indicative that survival remains associated with this type of reserve but determined by the previous generation of females. The decrease in WE between June and October could well illustrate the importance of the long generation time of G4. Early winter subadults and adults probably originate from the autumn G3 generation with relatively high wax esters percentage, while individuals found in late winter in August were probably new recruits from the winter G4 population with low WE but relatively high TAG percentages.
Fatty acid composition of D. pectinatus cannot be compared to any similar species from the Southern Ocean, since data is non-existent. Polar lipid structure showed a dominance of DHA, EPA and 16:0, a feature common to all marine planktonic crustaceans (Albers et al. Reference Albers, Kattner and Hagen1996, Mayzaud Reference Mayzaud1997, Lee et al. Reference Lee, Hagen and Kattner2006). However, the changes in PUFA proportion with season and generation illustrate the food limitation on the winter generation to ensure proper membrane structure. The lower content in DHA and 18:4n-3 together with the relatively stable level of EPA suggests limitation of DHA synthesis in relation to the very low level of food supply and the detritic nature of the particulate matter associated with the maximum of heterotrophic bacterial abundance (Razouls et al. Reference Razouls, De Bovee, Dellile, Fiala and Mayzaud1997).
Such a dietary constraint is expected to be even more influential on the control of the composition of neutral lipid. Indeed, triacylglycerols changes are indicative of the origin of food ingested over a relatively short time scale and follow quite well the changes in particulate characteristics (Razouls et al. Reference Razouls, De Bovee, Dellile, Fiala and Mayzaud1997). The spring and summer individuals were associated with markers of phytoplankton blooms: i.e. spring females with both diatoms and flagellates markers (PUFAs of the n-3 family) and summer ones with diatom (C16 PUFAs). Autumn and early winter individuals appeared to shift feeding towards a more omnivorous/carnivorous mode with C20 and C22 monoenes as characteristic descriptors (see Lee et al. Reference Lee, Hagen and Kattner2006). Late winter individuals are clearly associated with detritus/bacteria food chain associated with saturated and branched chain fatty acids. Wax esters illustrated a slightly different perspective related to the longer time integration of the fatty acid composition (Lee et al. Reference Lee, Hagen and Kattner2006). Hence, WE from the spring individuals are related to flagellate and some bacterial descriptors (18 PUFA, 16:4n-3, 20:3n-3 but also anteiso17:0, iso 17:0, 16:1n-5), while summer WE are clearly associated with diatom bloom markers (C16 PUFA and 20:5). Autumn and winter WE, associated with monoenes (C16, C20 and C22 as well as 16:3n-3, 20:4n-3 and 18:2n-6), would indicate a more omnivorous/carnivorous trophic mode.
The alcohol fraction is extremely stable in all generations with mainly saturated constituents accounting for more than 90%. The dominance of 16:0 and 14:0 alcohols is shared with other neritic copepods such as Pseudocalanus sp. (Fraser et al. Reference Fraser, Sargent and Gamble1989), Pseudocalanus acuspes (Norrbin et al. Reference Norrbin, Olsen and Tande1990) and Pseudocalanus minutus (Mayzaud Reference Mayzaud1980, Lischka & Hagen Reference Lischka and Hagen2006). Kattner et al. (Reference Kattner, Hagen, Falk-Petersen, Sargent and Henderson1996) suggested that such biosynthetic pathway is associated with omnivorous/carnivorous trophic behaviour. Given our results and the phytoplankton regime of late stages D. pectinatus in spring and summer, one could question this interpretation. Sargent et al. (Reference Sargent, Lee and Nevenzel1976) suggested that wax esters formation in copepods was a mechanism which ensures a high rate of lipid synthesis by converting inhibitory end products fatty acyl coenzyme-A into alcohols, which are removed from the system by being coupled with fatty acids of dietary origin. Fatty alcohols may originate either from phytoplankton fatty acids or synthesized de novo from carbohydrates or amino acids (Sargent & Henderson Reference Sargent and Henderson1986). In both cases D. pectinatus limits conversion to short chain alcohols, in contrast to large Calanus species, which convert monoenoic acids into 20:1n-9 and 22:1n-11 alcohols. All large calanoid species, which undergo diapause during winter to cope with low food conditions, accumulate wax esters over the short phytoplankton abundance with a dominance of high energy long chain C20 and C22 monoenoic acids (Falk-Petersen et al. Reference Falk-Petersen, Mayzaud, Kattner and Sargent2009). Small, neritic calanoid copepods, such as D. pectinatus, also feed mostly on phytoplankton during spring and summer and display a high rate of lipid synthesis. The outcome of these reserves seems unlikely to be oriented towards winter survival since few individuals from summer generation are likely to survive as long as winter. None of the small species mentioned are known to diapause and probably do not require such long chain high energy molecules for their metabolic processes. Other species with short chain saturated alcohols include herbivorous/omnivorous Calanus propinquus (Hagen et al. Reference Hagen, Kattner and Graeve1993), Metridia longa (Sargent & Henderson Reference Sargent and Henderson1986), Euchaeta japonica (Lee et al. Reference Lee, Nevenzel and Lewis1974) and carnivorous species such as Paraeuchaeta antarctica (Mayzaud unpublished data). Hence, long chain alcohols are more likely to be related to life cycle requirements associated with diapause than trophic type.
The changes in lipid classes over growth revealed some interesting features. Young D. pectinatus copepodites showed a dominance of TAG at stage C3, which is gradually replaced by wax esters from stage C4 to a maximum percentage during stage C5. Comparison is again difficult since most relevant data concerns large calanoid species but the increased content of wax esters with copepodite stage is consistent with the findings of Lee et al. (Reference Lee, Nevenzel and Lewis1974) on Euchaeta japonica, and of Kattner et al. (Reference Kattner, Graeve and Hagen1994) on Calanus propinquus, Calanoides acutus and Rhyncalanus gigas. The decrease in TAG seems more specific to such small species and may simply indicate a slow change from triacylglycerols dominating eggs and nauplii to low triglyceride/high wax esters composition in late copepodite and adult stages.
Fatty acid changes in relation to ontogenetic changes have been studied in different copepod species but generally with regard to total lipid rather than lipid classes or only for late copepodite stages (Lee et al. Reference Lee, Nevenzel and Lewis1974, Kattner & Krause Reference Kattner and Krause1987, Kattner et al. Reference Kattner, Graeve and Hagen1994, Ward et al. Reference Ward, Shreeve and Cripps1996). Hence, comparisons with the present findings are difficult due to the confounding effect of neutral lipid changes on the polar lipid composition. One of the interesting features recorded in the present work is the low percentages of EPA and DHA in the phospholipids of copepodite stage C3, and the recovery of high levels of DHA as early as stage C4 or EPA at stage C5. The same trend was observed for 18:4n-3 but over a smaller range of percentages. The reverse pattern could be seen for the n-6 PUFA suggesting a greater requirement in younger stages. Confirmation of this possibility requires further study on a more complete set of developmental stages. Saturated and monoenoic acids decreased with increasing growth stage.
Considering the limited ability to elongate and/or desaturate 18-carbon PUFA (Moreno et al. Reference Moreno, Moreno and Brenner1979), copepods require pre-formed essential fatty acids such as arachidonis acid (ARA, 20:4n-6), EPA and DHA in their diet to sustain growth. Although our study was not organized to study essential fatty acid (EFA) requirements, the data on the fatty acid partitioning between neutral and PL suggests useful information concerning EFA incorporation into membrane lipids. The proportions of DHA, EPA and to a lower extent ARA recorded in all stages (respectively 12–36%, 6–25% and 0.2–0.5%) were markedly higher than in the natural spring–summer particulate matter used as diet by the small copepods (DHA: 0–6%, EPA: 1–10%, ARA: 0.1–2%; Kerguelen shelf water, Mayzaud unpublished data), indicating selective incorporation into membrane lipids. This suggests that diet lipids are very probably limiting in terms of DHA and EPA incorporation mainly for the early C3 stage which feeds on the smaller fraction of the size particles with low PUFA content (Mayzaud et al. Reference Mayzaud, Chanut and Ackman1989).
In conclusion, while yearly and multiyear generation copepods have developed lipid accumulation to respond to winter survival and reproduction, multigenerational species such as Drepanopus appear to accumulate lipid for immediate specific needs, with spring and summer generation accumulation to ensure reproduction and autumn generation accumulation followed by winter catabolism to overcome late winter poor food conditions. The dominant control of neutral lipid composition by food intake, illustrated well the succession of trophic interaction imposed on each generation and the differential food sources for the different developmental stages. Assimilated and stored (n-3) as well as (n-6) PUFA are required to maintain structural synthesis, but further work on a larger set of developmental stages would be required to confirm our preliminary results.
Acknowledgements
The study was part of the IOZ programme financially supported by a grant from the Groupement de Recherche en Environnement 1069 “Ecosystèmes Polaires et Anthropisation” from the CNRS, from IPEV Interactions Oiseaux-Zooplancton 166 (IOZ), from Europe Noe MARBEF and by CNRS UMR 7093. Fieldwork was supported financially and logistically by the Institut Français pour la Recherche et la Technologie Polaires. The authors would like to thank Dr J. Dolan for his editorial comments and Dr F. Alonzo for his help during the fieldwork. The work would have been difficult without the help at sea of the captain and the crew from the RV La Curieuse. The technical help from P. Le Jeune, G. Roudaut and N. Coffineau was greatly appreciated.
