Abstract

Experiments were carried out to test the use of algal pigments in zooplankton grazing studies with a special emphasis on estimation of food selection. The results demonstrated that pigment composition of the phytoplankton food was reflected closely in the three copepod species Centropages typicus, Temora longicornis and Acartia tonsa, as well as in their faecal pellets. The fate of the phytoplankton pigments was studied in A. tonsa fed a diatom and a cryptophyte at a low and a high prey concentration. The concentration of gut pigments generally declined rapidly within the first 5–10 min after feeding terminated. The decline in pigment concentration was faster in copepods fed high concentrations of phytoplankton and specially when fed the diatom. However, after 3h of no feeding only minor changes in gut pigment composition were found mainly in alloxanthin, chlorophyll a and diadinoxanthin. Traditional grazing experiments were carried out in parallel with pigment analysis in experiments where the copepod A. tonsa was exposed to a mixture of food organisms. The results demonstrated that the two methods gave similar results with regard to food selection and that with certain precautions, pigment analysis can be successfully used in food selection studies.

INTRODUCTION

Copepods are along with microzooplankton the major consumers of the primary production in the oceans and make up a very important link between the phytoplankton production and upper tropic levels such as fish. To understand the function of the pelagic ecosystem and the energy flow between the different trophic levels, it is important to be able to measure feeding rates and any food selectivity of the herbivorous zooplankton.

The feeding rate of marine zooplankton in their natural environment has traditionally been estimated using the gut fluorescence technique that quantifies the amount of fluorescent compounds in the animal gut, where chlorophyll a (Chl a) is the most widely used proxy for phytoplankton biomass. However, inconsistent degradation of Chl a complicates quantitative interpretations based on this method. In Acartia clausi Chl a degradation varied between 30 and 94% depending on the food concentration of Thalassiosira weissflogii and the loss was positively related to total ingestion (Tirelli and Mayzaud, 1998). The quantity of pigment lost during gut passage of this marine copepod was directly related to total ingestion, and Chl a degradation was higher for copepods that had been acclimated at high food concentration. Thus, degradation of ingested Chl a may depend on food concentration and ingestion rate within the individual species and it furthermore differs between species (Penry and Frost, 1991; Tirelli and Mayzaud, 1998).

Copepods can be selective in the food they consume (Head and Harris, 1994; Meyer-Harms and von Bodungen, 1997; Meyer-Harms et al., 1999), and therefore the phytoplankton community composition is not necessarily reflected in the food ingested by the copepods. In addition to Chl a, that is found in all phytoplankton and is the basis of the gut fluorescence technique, phytoplankton contains a range of more-or-less group-specific other chlorophylls (e.g. Chl b, Chl c1, Chl c2, Chl c3) and carotenoid pigments (e.g. astaxanthin, α- and β-carotenes) (Jeffrey et al., 1997). Thus, phytoplankton composition can be estimated from analyses of characteristic pigments by high-performance liquid chromatography (HPLC) and subsequent data analysis using the matrix factorization program CHEMTAX (Mackey et al., 1996). HPLC analyses allow us to include numerous marker pigments in estimating copepod food selection and several studies have used pigments to study food selectivity by zooplankton (Quiblier-Llobéras et al., 1996; Meyer-Harms et al., 1999; Descy et al., 1999; Wong et al., 2006). Results from these studies are, however, somewhat contradictory. Comparisons of selective feeding determined from the pigment method with measurements of cell removal have shown similar selection results for the marine calanoid copepod Acartia bifilosa (Meyer-Harms and von Bodungen, 1997), the marine cladoceran Penilia avirostris (Wong et al., 2006) and the freshwater cladoceran Daphnia galeata (Thys et al., 2003). In contrast, Pandolfini et al. (Pandolfini et al., 2000) showed that gut content of pigments reflected qualitatively the type of phytoplankton ingested by D. galeata while pigments found in the freshwater copepod Eudiaptomus gracilis differed from the diet indicating different resilience of pigments in different genera of herbivorous zooplankton. Kleppel et al. (Kleppel et al., 1988) found differences in the presence of the marker pigment fucoxanthin between two species of marine copepods (Calanus and Clausocalanus) and differences related to food concentration in one of these species. In the freshwater copepod Diaptomus minutus, the ingested carotenoid alloxanthin registered in the gut while ingested fucoxanthin was not found (Descy et al., 1999). Thus, the different marker pigments are processed differently during gut passage of different herbivores.

If pigments are to be used to investigate food selection in marine copepods in the field it is essential to know how to detect and separate pigments in the copepods, how long the pigments remain detectable in the copepods after they have been ingested, if they break down or are transformed to other products, if breakdown differs among pigments and/or among copepod species and the effects of gut passage time on ingested pigments. We tested these requirements using three tests; firstly by testing if pigment composition from the cryptophyte Rhodomonas salina and the diatom T. weissflogii could be recovered in the marine copepods Acartia tonsa, Centropages typicus or Temora longicornis and their faecal pellets and, secondly by observing how long pigments could be detected in A. tonsa after the copepods were exposed to no food and if degradation patterns were similar for all detected pigments during a 3-h period. Finally, an experiment was conducted to compare if the HPLC and cell count methods showed the same food selection by A. tonsa feeding on a mixture of R. salina and T. weissflogii.

METHOD

Phytoplankton cultures

The cryptophyte R. salina and the diatom T. weissflogii were used as food for the copepods in our experiments. Both cultures were grown in 32‰ pasteurized seawater enriched with B1 nutrient solution (Hansen, 1989). Sodium silicate was added to the diatom culture. The algal cultures were grown at 17.5 ± 1.5°C in a 16:8 h light:dark cycle at an irradiance of 100 µmol photons m−2 s−1. Average algal cell volumes and all phytoplankton counts were measured using a Beckman Coulter Multisizer™ 3 Coulter Counter® with a calibrated 100-mm orifice tube. Cell concentrations were converted to carbon using the volume-based estimates from Menden-Deuer and Lessard (Menden-Deuer and Lessard, 2000) and Strathmann (Strathmann, 1967).

Copepod cultures

Three species of copepods were used for the experiments: C. typicus, T. longicornis and A. tonsa. All species were isolated from the North Sea and maintained in cultures at DTU-Aqua for over a year. Before the experiments the copepods were well fed with either a pure diet of R. salina (A. tonsa) or a mixed diet of R. salina supplemented with T. weissflogii (C. typicus and T. longicornis). Copepod cultures were maintained at 17.5 ± 1.5°C. Copepods used in the experiments had hatched from eggs 4–5 weeks prior to the experiments.

Experiment 1: fate of pigments during feeding

The fate of the algal pigments after ingestion and defecation was studied by feeding the three copepod species, C. typicus, T. longicornis and A. tonsa either R. salina or T. weissflogii. One hundred adult copepods were put in each of three replicate 5 L buckets containing 400 µg C L−1 of the respective food suspension. Each copepod species was acclimated to their respective diets for 24 h prior to the experiments. After acclimation, the copepods were transferred to new food suspensions and allowed to feed for 48 h. The same setup, but with copepods in filtered seawater was run as a control. At the end of all the experiments, the triplicate samples of 100 individuals each were placed on a 25-mm GF/C filter and wrapped in aluminium foil and immediately frozen at −80°C. Care was taken to process samples quickly, and not to expose the samples to light and to minimize possible pigment degradation.

To collect faecal pellets, ∼300 copepods were put in a 5-L tank with either R. salina or T. weissflogii at ∼400 µg C L−1. After 48 h of grazing, faecal pellets were collected from the bottom of the tanks, by sucking them very gently using a long tube. Algal prey was washed very gently from the pellets on the sieve (mesh size 45 µm) with sterile seawater to avoid contamination. Pellets were then washed onto a 25-mm GF/C filter, wrapped in aluminium foil and immediately frozen at −80°C.

Experiment 2: gut clearance of pigments in copepods

Gut clearance time in A. tonsa females was measured using both R. salina and T. weissflogii at two different concentrations: 40 and 400 µg C L−1. Prior to the experiment the copepods had been acclimated to the specific algae and concentrations for 24 h. Fifty female A. tonsa were added to each of 33 glass beakers (600 mL) containing the experimental food concentration and a cylinder with a 200 -µm mesh net fitted to the bottom and suspended in the beaker. The copepods were placed inside the cylinders to prevent them from feeding on their own faecal pellets as the pellets sank out of the cylinder. The copepods were allowed to feed for 24 h. This ensured that their guts were full at the start of the experiment. In addition, at the start of the gut evacuation measurements, this system permitted a fast transfer of females with minimum stress from the food medium to a beaker filled with clean filtered seawater. The experiment consisted of three replicate trial measurements of gut content. Samples were taken at 11 time intervals: 0, 5, 10, 15, 20, 30, 45, 60, 90,120 and 180 min after transfer. At the time of the first sampling, (full gut) the first triplicate samples of copepods inside the cylinders were gently transferred from the food suspension to a bowl of filtered seawater to quickly rinse algae out before the copepods were washed onto a 25-mm GF/C filter. At the successive sampling times the copepods were washed directly onto the filters.

Experiment 3: use of pigments to estimate food selection

Acartia tonsa was fed a mixed diet of R. salina and T. weissflogii in the ratio of 1:2 (carbon units) at a total concentration of 730 µg C L−1. Experiments were carried out at an irradiance of 10 µmol photons m−2 s−1 and algal cultures were kept at this irradiance for 24 h prior to the experiments. Nutrients were added to all experimental bottles to prevent differentiation between grazing bottles and controls due to NH4+ enrichment from animals. Copepods were acclimated for 24 h prior to the experiment to the food concentration and irradiance. Fifteen A. tonsa females were added to each of the triplicate 630-mL experimental bottles with food suspensions. The bottles were sealed without head space and placed on a rotating plankton wheel (1 rpm) and incubated for ∼24 h. In addition to the experimental bottles, five control bottles without animals were incubated under the same conditions. At the termination of the experiment cell concentrations in each bottle were counted and ingestion rates calculated according to Frost (Frost, 1972). From bottles 1–4, 5–6 and 7–9 copepods were pooled onto three 25-mm GF/C filters and wrapped in aluminium foil and immediately frozen at −80°C. A 50 mL water sample from each bottle was filtered through a 25-mm GF/C filter which was wrapped in aluminium foil and immediately frozen at −80°C. These samples were subsequently analysed for pigments with HPLC.

Preparation of samples for HPLC analyses

Samples of the laboratory cultures of phytoplankton and samples of copepods were prepared for HPLC analyses by two different methods. Every step of the preparation, from the −80°C aluminium foil wrapped samples to the vials containing the pigment extraction was done at very low light intensity and very quickly to minimize pigment degradation.

GF/C filters with copepods were unwrapped and transferred to glass vials. One millilitre acetone was added and the contents were sonicated (Sonics Vibra Cell ultrasonic processor, Newtown, CT, USA) for ∼10 s and placed in the dark at 4°C for 24 h. The extracts were filtered (0.2 µm) into an HPLC vial and placed in the HPLC auto sampler. Filters with phytoplankton were unwrapped and placed in a 10 mL syringe with a 0.2-µm filter placed at the tip of the syringe. After addition of 2.5 mL acetone the content was sonicated for 10 s and pushed through the filter into a glass vial. One millilitre of the filtered extract was transferred to an HPLC vial where 0.25 mL distilled water was added prior to the analysis.

HPLC analyses were performed on a Shimazu LC 10A system with a Supelcosil C18 column (250 × 4.6 mm, 5 µm) using the method of Wright et al. (Wright et al., 1991). Pigments were identified by retention times and absorption spectra identical to those of authentic standards, and quantified against commercial standards (DHI Water & Environment, Hørsholm, Denmark). Monadoxanthins and crocoxanthins concentration were tentatively quantified using the HPLC response factor for lutein with a similar absorption spectrum and retention time. Free astaxanthin and its esters were identified and quantified by retention time/absorption spectrum and HPLC response factor, respectively, of commercially available astaxanthin (Sigma™). Phaeopigments were detected at 665 nm wavelength while all other pigments were detected at 436 nm.

The HPLC method used did not separate Chl c1 and Chl c2. Thus data on Chl c may represent a combination of these two chlorophylls.

Statistics

All statistical analyses were performed using the software SigmaStat 3.1. Differences between measured variables were tested for significance using a one-way ANOVA or Kruskal–Wallis one-way analysis of variance on ranks if the data did not meet the requirement of normality or equality of variance. Both tests were followed by the Holm–Sidak method or Dunn's method for post hoc comparison. For the food selection experiment t-tests were made using SigmaStat 3.5.

RESULTS

Experiment 1: fate of pigments during feeding

Rhodomonas salina

The pigment profile for R. salina consisted of mainly Chl a (57%) and alloxanthin (22%) with additional minor contributions from Chl c (8%), α-carotene (5%), monadoxanthin (4%) and crocoxanthin (4%; Table 1). The pigment profile of all the three copepod species fed R. salina was very similar to the profile of R. salina (Fig. 1a) with the exception of monadoxanthin that was not detected in the copepods. A significant difference was found in the relative abundance of Chl c (lower than in R. salina, one-way ANOVA, df = 3, P = 0.011) and Chl a (higher than in R. salina, one-way ANOVA, df = 3, P = 0.008) in all copepods and alloxanthin that was significantly lower in A. tonsa compared with the diet (Holm–Sidak, post hoc comparison, P = 0.007). There were no significant differences between the copepods in their relative pigment composition feeding on R. salina, with the exception of the pigment crocoxanthin between T. longicornis and A. tonsa (Dunn's, post hoc method, P < 0.05) and α-carotene which was only found in A. tonsa.

Table I:

Weight fraction of pigments (% ± 1 SE) in Rhodomonas salina and Thalassiosira weissflogii

 R. salina T. weissflogii 
Pigments % ± SE % ± SE 
Chlorophyll c 7.9 ± 2.0 3.3 ± 0.3 
Fucoxanthin nd 24.5 ± 1.0 
Diadinoxanthin nd 6.5 ± 0.6 
Alloxanthin 22.1 ± 1.2 nd 
Monadoxanthin 3.9 ± 1.3 nd 
Crocoxanthin 4.4 ± 0.5 nd 
Chlorophyll a 56.9 ± 5.4 62.9 ± 1.4 
α-Carotene 4.8 ± 0.5 nd 
β,β-Carotene nd 2.8 ± 0.02 
 R. salina T. weissflogii 
Pigments % ± SE % ± SE 
Chlorophyll c 7.9 ± 2.0 3.3 ± 0.3 
Fucoxanthin nd 24.5 ± 1.0 
Diadinoxanthin nd 6.5 ± 0.6 
Alloxanthin 22.1 ± 1.2 nd 
Monadoxanthin 3.9 ± 1.3 nd 
Crocoxanthin 4.4 ± 0.5 nd 
Chlorophyll a 56.9 ± 5.4 62.9 ± 1.4 
α-Carotene 4.8 ± 0.5 nd 
β,β-Carotene nd 2.8 ± 0.02 

nd, not detectable.

Fig. 1.

Temora longicornis (TL), Acartia tonsa (AT) and Centropages typicus (CT). Relative pigment composition of phytoplankton, animals and their faecal pellets, after feeding on (a) Rhodomonas salina (RS) and (b) Thalassiosira weissflogii (TW). Whiskers indicate 1 SE.

Fig. 1.

Temora longicornis (TL), Acartia tonsa (AT) and Centropages typicus (CT). Relative pigment composition of phytoplankton, animals and their faecal pellets, after feeding on (a) Rhodomonas salina (RS) and (b) Thalassiosira weissflogii (TW). Whiskers indicate 1 SE.

The relative pigment composition in the faecal pellets was neither different between the copepod species nor from the composition found in the phytoplankton (Holm–Sidak, post hoc comparison, P > 0.05). The only exception was C. typicus faecal pellets that had significantly higher alloxanthin concentration compared with T. longicornis and A. tonsa faecal pellets (Holm–Sidak, post hoc comparison, both P = 0.002).

Thalassiosira weissflogii

The pigment profile of T. weissflogii consisted mainly of Chl a (63%) and fucoxanthin (24%) with additional minor contributions from diadinoxanthin (6%), Chl c (3%) and β,β-carotene (3%; Table 1). The pigment composition in the three copepods was very similar to that found in the phytoplankton food source (Fig. 1b) (one-way ANOVA, P = 0.099, P = 0.288 and P = 0.514 for Chl c, fucoxanthin and diadinoxanthin, respectively). The only copepod retaining β,β-carotene was A. tonsa.

The pigment composition of A. tonsa and C. typicus faecal pellets did not differ from their T. weissflogii diet. However, T. longicornis faecal pellets differed significantly in both fucoxanthin and diadinoxanthin content compared with T. weissflogii (Holm–Sidak, post hoc comparison, P < 0.05 and P = 0.008, respectively). The fucoxanthin content of T. longicornis faecal pellets was significantly lower compared with both A. tonsa and C. typicus faecal pellets (Holm–Sidak, post hoc comparison, P < 0.05).

Some pigments could not be identified in the copepods due to overlapping retention times with the copepod pigments astaxanthin and astaxanthin esters. This problem was evident for monadoxanthin that was impossible to separate from the high content of astaxanthin present in the copepods, and α-carotene and β,β-carotene that overlapped with different astaxanthin esters present in high amounts in T. longicornis and C. typicus. Acartia tonsa was the only one of the three copepod species tested that did not have high amounts of astaxanthin esters which was the reason for selecting this copepod for further experiments on food selectivity.

Experiment 2: copepod gut clearance

After feeding on R. salina at a food concentration of 40 µg C L−1 no decrease was detected in total gut pigment concentrations of A. tonsa over 3h of non-feeding (t-test P = 0.341 and P= 0.209 for 0–5 and 0–10 min, respectively). No significant differences were observed in any of the specific gut pigment concentrations after 3h (one-way ANOVA, Holm–Sidak, Fig. 2a) and thus the ratio of pigments in the gut did not change over time (Fig. 2c). In the case, where A. tonsa had been feeding in the 400 µg C L−1R. salina suspension prior to the experiment, the total gut pigment concentrations decreased rapidly over the first 5 min, stabilizing after 5–10 min of no feeding (Fig. 2b). There was a significant difference in Chl a concentration between 5 and 180 min (t-test, P = 0.013). However, the decrease in alloxanthin was not significant for the first 20 min but first after 30 min (t-tests, P = 0.045). There were no significant differences in the concentration for any of the other pigments during the 3h in filtered seawater. The changes in the pigment concentrations were only reflected in the relative pigment composition with the drop in alloxanthin in the gut after 30 min, after which it remained stable (Fig. 2d).

Fig. 2.

Acartia tonsa. Concentration of gut pigments (ng individual−1 ± 1SE) with time after feeding was stopped after feeding on food solution of (a) 40 µg C L−1Rhodomonas salina and (b) 400 µg C L−1R. salina. Relative contribution of the pigments in the gut after feeding (c) 40 µg C L−1R. salina and (d) 400 µg C L−1R. salina. Right-hand axis in (b) represents the concentration scale for Chl a.

Fig. 2.

Acartia tonsa. Concentration of gut pigments (ng individual−1 ± 1SE) with time after feeding was stopped after feeding on food solution of (a) 40 µg C L−1Rhodomonas salina and (b) 400 µg C L−1R. salina. Relative contribution of the pigments in the gut after feeding (c) 40 µg C L−1R. salina and (d) 400 µg C L−1R. salina. Right-hand axis in (b) represents the concentration scale for Chl a.

When A. tonsa had been fed T. weissflogii at a food concentration of 40 µg C L−1 prior to starvation there was no significant difference in concentrations of fucoxanthin or Chl a over time (Kruskal–Wallis one-way ANOVA on ranks, P = 0.052 and P = 0.051, respectively, Fig. 3a). Diadinoxanthin was found only in the initial samples after which it was no longer detectable. A significant difference over time was found for β,β-carotene (Kruskal–Wallis one-way ANOVA on ranks, P = 0.027). The relative pigment composition in the gut did not reflect these rapid concentration changes, with the exception of the presence of diadinoxanthin in the initial sample that reduced the relative contribution of Chl a (Fig. 3c). In the cases where A. tonsa had been fed T. weissflogii at a prey concentration of 400 µg C L−1 prior to starvation, significant differences in concentrations of all pigments but Chl c was found within the first 5 min (t-tests: Chl a P = 0.014, fucoxanthin P = 0.011, diadinoxanthin P = 0.015, β,β-carotene P = 0.020 and Chl c P = 0.183, Fig. 3b). One-way ANOVA analysis of samples taken between 5 min and 3 h showed that there were no significant changes in any of the pigments after the first 5 min. The relative contribution of the pigments to the gut composition was stable for 120 min when diadinoxanthin was not longer detectable (Fig. 3d). The contribution of fucoxanthin decreased slightly with time.

Fig. 3.

Acartia tonsa. Concentration of gut pigments (ng individual−1 ± 1SE) with time after feeding was stopped on food solution of (a) 40 µg C L−1Thalassiosira weissflogii and (b) 400 µg C L−1T. weissflogii. Relative contribution of the pigments in the gut after feeding (c) 40 µg C L−1T. weissflogii and (d) 400 µg C L−1T. weissflogii. Note the different scales on the axes and the right hand axis in (b) that represents the concentration scale for Chl a, fucoxanthin and diadinoxanthin. Error bars are sometimes smaller than symbols and not visible.

Fig. 3.

Acartia tonsa. Concentration of gut pigments (ng individual−1 ± 1SE) with time after feeding was stopped on food solution of (a) 40 µg C L−1Thalassiosira weissflogii and (b) 400 µg C L−1T. weissflogii. Relative contribution of the pigments in the gut after feeding (c) 40 µg C L−1T. weissflogii and (d) 400 µg C L−1T. weissflogii. Note the different scales on the axes and the right hand axis in (b) that represents the concentration scale for Chl a, fucoxanthin and diadinoxanthin. Error bars are sometimes smaller than symbols and not visible.

Experiment 3: use of pigments to estimate food selection

Cell counts showed that the ingestion rate of A. tonsa on R. salina was 0.95 ± 0.06 µg C animal−1 d−1 (±1 SE) and 4.85 ± 0.20 µg C animal−1 d−1 while feeding on T. weissflogii. The selection indices (α, Chesson, 1983) calculated were α = 0.73 ± 0.03 for T. weissflogii and α = 0.27 ± 0.03 for R. salina. There was a significant positive feeding selectivity for T. weissflogii (t-test, P < 0.001). HPLC analyses of the pigment content in the copepods after grazing on the mixed food showed that they contained T. weissflogii and R. salina in the proportion 0.77:0.23 based on α- and β,β-carotene content of the copepods and 0.66:0.34 based on fucoxanthin and alloxanthin content of the copepods (Fig. 4). Kruskal–Wallis one-way ANOVA on the ratios between proportions based on α- and β,β-carotene content of the copepods and the selection index α based on cell counts was conducted to test if the pigment-based proportions in the copepods were the same as the calculated ingested selection ratios. No significant difference was found between the two methods (P = 0.133). However, the ratios between proportion based on alloxanthin and fucoxanthin content of the copepods and selection index α showed that there was a significant difference between the methods (one-way ANOVA P = 0.029). No difference between counts and the combined specific pigment ratios α-carotine and alloxanthin (Rho) versus β,β-carotene and fucoxanthin (TW) was found (t-test, P = 0.626).

Fig. 4.

Acartia tonsa. Selection index α for mixture of Rhodomonas salina (grey bars) and Thalassiosira weissflogii (black bars) based on cell removal, and the ratio between the species-specific pigments alloxanthin and α-carotene (Rho) and fucoxanthin and β-carotene (TW) in the gut. Whiskers are +1 SE.

Fig. 4.

Acartia tonsa. Selection index α for mixture of Rhodomonas salina (grey bars) and Thalassiosira weissflogii (black bars) based on cell removal, and the ratio between the species-specific pigments alloxanthin and α-carotene (Rho) and fucoxanthin and β-carotene (TW) in the gut. Whiskers are +1 SE.

DISCUSSION

Estimation of immediate feeding and food selection in natural zooplankton populations is important when studying vital rates in nature. However, it has proved to be a great challenge to find a suitable method to use and most often feeding is not measured, resulting in black box studies on various vital rates. The best method to use is to directly measure cell removal with bottle incubations, but this method is labour intensive and requires time consuming counting and therefore not often used in nature where food can be very diverse. The frequently used gut fluorescence method has proved to be problematic due to rapid pigment degradation (see above). However, the use of specific phytoplankton pigments to estimate immediate feeding and selection by the copepod community is an attractive but not well-tested method.

To make accurate estimates of food selection using the gut pigment approach, the pigments should be detectable in the organism of interest in the same proportion as the representative food type ingested. In all the three copepod species studied the overall composition of phytoplankton pigments in the gut was very similar to that of the food. This agrees with some studies, but contrasts with others.

The fate of carotenoids, the main phytoplankton marker pigments, upon ingestion by zooplankton has been studied before in both freshwater and marine species but with varying results. Pandolfini et al. (Pandolfini et al., 2000) showed that pigment processing during gut passage can be different between zooplankton species/groups (e.g. a copepod vs. a cladoceran) while our study and the studies of Quiblier et al. (Quiblier et al., 1994) and Quiblier-Llobéras et al. (Quiblier-Llobéras et al., 1996) did not find any significant differences between the zooplankton species tested. Pandolfini et al. (Pandolfini et al., 2000) found almost exclusively alloxanthin in the zooplankters where other pigments were lost in high proportions and they concluded that for their two species data on gut pigments should be used with caution. Similar results on the presence of alloxanthin but poor recovery of diadinoxanthin and fucoxanthin in the gut were obtained for D. minutus (Descy et al., 1999). In all three copepods tested in our study, the pigments appeared in the gut in similar ratios as in the food (Fig. 1).

Also in contrast to our study, other studies have found that some pigments from the diet were not present in the faecal pellets. Fucoxanthin and 19′-hexanoyloxyfucoxanthin did not appear in the faecal pellets from the marine copepod Calanus spp. after they had been grazing on a diet dominated by Chl a, Chl c, fucoxanthin and 19′-hexanoyloxyfucoxanthin (Head and Harris, 1994) and fucoxanthin and diadinoxanthin from the diet appeared only as traces in the faecal pellets from the freshwater copepod D. minutus (Descy et al., 1999). Such pigment loss was not apparent in the faecal pellets in the present study, where a close similarity was found between the pigment compositions in copepods, their faecal pellets and the food they had been eating (Fig. 1).

It is difficult to evaluate the reasons for the different results in the present and the above mentioned studies. Several factors have been found to affect pigment degradation; light (Head, 1992), previous feeding history (Penry and Frost, 1991; Head, 1992), food concentration (Kleppel et al., 1985; Head and Harris, 1992, 1994) and phytoplankton species composition (Head, 1992; Quiblier et al., 1994). Even though it is not always specifically mentioned, we expect that care has been taken with light in all the previous studies. However, all the other factors may have varied between studies. An important difference may have been the food concentrations and availability that is then related to copepod concentration in the experiments. In an attempt to compare the results of the studies that give information on food availability and zooplankton densities, we made a rough calculation of food availability (assuming a C/chl ratio of 50) and predator densities using a best estimate of zooplankton carbon content, based on literature values for the species used in the studies. This gives a carbon value for food per predator carbon; Cfood:Czoo (Table 2). Carbon content (µg) of Calanus spp. is based on carbon regressions from Swalethorp et al. (Swalethorp et al., 2011) and of the smaller copepods based on the estimate logC = 2.5 logPL − 7.5 (based on Table 35 in Mauchline, 1998), where PL is an estimate of prosome length (µm) of the predators used. The calculation suggests (Table 2) that studies showing the highest pigment degradation have very low food availability per animal (<0.05 µgCfood µgCzoo−1; Head and Harris, 1992, 1994; Pandolfini et al., 2000), while studies with higher concentrations (over 3 µgC µgC−1) show lesser pigment destruction (Meyer-Harms and von Bodungen, 1997; this study). The study of Descy et al. (Descy et al., 1999), however, does not fit into this pattern. At the low food concentrations (40 µg C L−1) in our study, pigment recovery was poorer than at higher concentrations. Alloxanthin and Chl c did not appear in the A. tonsa gut while feeding on the low concentration of R. salina, and Chl c was missing in the copepod gut when fed on the low concentration of T. weissflogii (compare diets in Fig. 1 to time 0 in Figs 3 and 4). As the predator–prey ratio can vary greatly in the field depending on season, this should be kept in mind when using gut pigments as an estimate for food selection, as well as when deciding predator concentrations during experiments.

Table II:

Comparison of selected studies reporting on the fate of different pigments in zooplankton gut and faecal pellets

Zooplankton genera ind. L−1 Food conc (µg C L−1Food type Cfood:Czoo All Dia Mon Fuc Hex Cha Chc αcar βcar Pra Per Zea Cro Lut References 
Gut 
 Acartia 55 450–1100 Natural 3.8–9.4 ++ nm nm nm nm nm nm nm ++ −− nm nm Meyer-Harms and von Bodungen (1997
 Diaptomus 63–95 2000–3000 Natural 2.8 − −− − nm nm nm nm nm nm nm −− Descy et al. (1999
 Eudiaptomus 200–280 180–575 Natural 0.1–0.3 ++ nm nm −− nm − nm nm nm nm nm nm nm −− Pandolfini et al. (2000
 Daphnia 200 460–1790  0.8–3.0 ++   −−          −− 
 Acartia 20 400 Rb 3.1 ++ −− nm ++ −− ++ nm nm nm nm This study Exp 1 
 Temora 20   1.6 ++  −−   ++ −− −−     ++  
 Centropages 20   1.7 ++  −−   ++ −− −−     ++  
 Acartia 20 400 TW 3.1 ++ ++ nm ++ ++ nm nm nm nm 
 Temora 20   1.6  ++  ++  ++ ++  −−      
 Centropages 20   1.7  ++  ++  ++ ++  −−      
 Acartia 100 40 Rb 0.1 −− −− nm −− nm nm nm nm This study Exp 2 
  TW  ++ ++  ++ −− ++     
 Acartia 100 400 Rb 0.6 ++ −− nm ++ −− ++ nm nm nm nm This study Exp 2 
  TW  ++ ++  ++ ++     
Faecal pellets 
 Eudiaptomus 200–280 180–575 Natural 0.1–0.3 ++ − nm −− nm − nm nm nm nm nm nm nm −− Pandolfini et al. (2000
 Daphnia 200 460–1790  0.8–3.0 −−   −          − 
 Diaptomus 63–95 2000–3000 Natural 2.8 ++ − nm − nm − nm nm nm nm nm nm nm −− Descy et al. (1999
 Calanus 30–45 100 Natural 0.01 nm − nm − nm −− −− nm nm nm nm nm nm nm Head and Harris (1992
 500  0.04    −−        
Mix of copepods 133 20 Natural 0.00 nm nm nm −− −− −− nm − − nm nm nm nm nm Head and Harris (1994
432 110  0.00      −         
 Acartia 20 400 Rb 3.1 ++ ++ nm ++ ++ nm nm nm ++ nm This study Exp 1 
 Temora 20   1.6               
 Centropages 20  TW 1.7 ++ ++  ++ ++ ++     
Zooplankton genera ind. L−1 Food conc (µg C L−1Food type Cfood:Czoo All Dia Mon Fuc Hex Cha Chc αcar βcar Pra Per Zea Cro Lut References 
Gut 
 Acartia 55 450–1100 Natural 3.8–9.4 ++ nm nm nm nm nm nm nm ++ −− nm nm Meyer-Harms and von Bodungen (1997
 Diaptomus 63–95 2000–3000 Natural 2.8 − −− − nm nm nm nm nm nm nm −− Descy et al. (1999
 Eudiaptomus 200–280 180–575 Natural 0.1–0.3 ++ nm nm −− nm − nm nm nm nm nm nm nm −− Pandolfini et al. (2000
 Daphnia 200 460–1790  0.8–3.0 ++   −−          −− 
 Acartia 20 400 Rb 3.1 ++ −− nm ++ −− ++ nm nm nm nm This study Exp 1 
 Temora 20   1.6 ++  −−   ++ −− −−     ++  
 Centropages 20   1.7 ++  −−   ++ −− −−     ++  
 Acartia 20 400 TW 3.1 ++ ++ nm ++ ++ nm nm nm nm 
 Temora 20   1.6  ++  ++  ++ ++  −−      
 Centropages 20   1.7  ++  ++  ++ ++  −−      
 Acartia 100 40 Rb 0.1 −− −− nm −− nm nm nm nm This study Exp 2 
  TW  ++ ++  ++ −− ++     
 Acartia 100 400 Rb 0.6 ++ −− nm ++ −− ++ nm nm nm nm This study Exp 2 
  TW  ++ ++  ++ ++     
Faecal pellets 
 Eudiaptomus 200–280 180–575 Natural 0.1–0.3 ++ − nm −− nm − nm nm nm nm nm nm nm −− Pandolfini et al. (2000
 Daphnia 200 460–1790  0.8–3.0 −−   −          − 
 Diaptomus 63–95 2000–3000 Natural 2.8 ++ − nm − nm − nm nm nm nm nm nm nm −− Descy et al. (1999
 Calanus 30–45 100 Natural 0.01 nm − nm − nm −− −− nm nm nm nm nm nm nm Head and Harris (1992
 500  0.04    −−        
Mix of copepods 133 20 Natural 0.00 nm nm nm −− −− −− nm − − nm nm nm nm nm Head and Harris (1994
432 110  0.00      −         
 Acartia 20 400 Rb 3.1 ++ ++ nm ++ ++ nm nm nm ++ nm This study Exp 1 
 Temora 20   1.6               
 Centropages 20  TW 1.7 ++ ++  ++ ++ ++     

Cfood:Czoo, food carbon per zoooplantkon carbon using carbon chlorophyll conversion factor of 50; All, alloxanthin; Dia, diadinoxanthin; Mon, monadoxanthin; Fuc, fucoxanthin; Hex, 19′-hexanoyloxyfucoxanthin; Cha, Chlorophyll a; Chc, Chlorophyll c; αcar, α-carotene; βcar, β,β-carotene; Pra, prasinoxanthin; Per, peridinin; Zea, zeaxanthin; Cro, crocoxanthin; Lut, Lutein; Rb, Rhodomonas baltica; TW, Thalassiosira weissflogii. Pigment recovered: ++: >75%; +: 50–75%; −:25–50%; − −: <25% compared with pigments in the food. / : not in the food suspension; nm: not measured.

The effect of food concentration on phytoplankton pigment concentration in A. tonsa showed that the pigment gut evacuation rate varied depending on pigments and food type. Gut evacuation was much faster when feeding on T. weissflogii compared with R. salina (Figs 2 and 3b) at 400 µg C L−1. If we take 0.3 ng pigment individual−1 as a baseline for a near-empty gut the ingestion on 40 µg C L−1R. salina was not significant for A. tonsa. This near-empty gut value was reached at 5 and 10 min for T. weissflogii at 400 and 40 µg C L−1, respectively, and first after 3 h for R. salina at 400 µg C L−1 (note different scale for Chl a). While it agrees with previous studies that showed that gut transit time was faster when copepods were feeding on diatoms compared with heterotrophic and autotrophic dinoflagellates, chlorophytes and a ciliate (Besiktepe and Dam, 2002), this rapid depletion of the Chl a pigment in the gut for the high 400 µg C L−1 TW concentration is more rapid than previously reported. Besiktepe and Dam (Besiktepe and Dam, 2002) showed gut passage time under 10 min at 250 µg C L−1 for A. tonsa. In our experiment the copepods were acclimated to supersaturated concentration for 48 h before the initial measurement and therefore started with completely full guts; 6 ± 1.4 ng Chl aind−1 (n= 3). This is a higher initial concentration than reported for A. tonsa by Tirelli and Mayzaud (Tirelli and Mayzaud, 2005) with 0.5 ng pigment individual−1 after 3 h acclimation at ∼250 µg C L−1. Gut transit time is faster when feeding on high prey concentrations compared with low (Besiktepe and Dam, 2002; Tirelli and Mayzaud, 2005). It has also been shown that chlorophyll a degradation is two to four times faster when acclimated/feeding at high food concentrations for a few days compared with low food levels (Penry and Frost, 1991; Tirelli and Mayzaud, 1998). Our rapid decline in Chl a pigment could therefore be a combination of these two factors. The pigment reduction was ∼17% for alloxanthin and 40% for Chl a during the first 5 min of R. salina (Fig. 2b), while gut evacuation was fast for all pigments when fed T. weissflogii (86–100% within the first 5 min). Interestingly, the species-specific pigments were still detectable in approximately the same ratio as in the respective food after 30–60 min of no food intake at the higher food concentration. Thys et al. (Thys et al., 2003) also found high amounts of pigments in freshwater cladocerans after 2 h of starvation in filtered water but related this to increased gut passage time due to low previous food levels.

One of the aims of the present study was to test if gut pigment analysis could be used for detection of food selection irrespective of pigment destruction in the gut with time. The results show that the method gave results comparable with those of traditional grazing experiments in estimating food selection from a mixed diet of two species of phytoplankton. The Chesson's selection indices α calculated from cell counts (0.73:0.27) of T. weissflogii and R. salina were within the range obtained from two different sets of characteristic pigments (0.66:0.34 and 0.77:0.23 for fucoxanthin:alloxanthin and β,β-carotene:α-carotene, respectively), indicating that the pigment approach can provide a reasonable estimate of food selection by A. tonsa when feeding on a mixture of these two species. However, it is entirely dependent on which pigments are chosen for detecting the phytoplankton groups. In our study some pigments appear to be more robust than others and the choice of pigments to use should be carefully evaluated (see also Meyer-Harms and von Bodungen, 1997).

In some studies the matrix factorization program CHEMTAX (Mackey et al., 1996) has been used to estimate what natural populations of copepods or cladocerans ingested (Meyer-Harms et al., 1999; Thys et al., 2003). The use of CHEMTAX is based on an input of ratios between the carotenoid pigments (and other chlorophylls) and Chl a in the different groups of phytoplankton potentially present. If applied to gut content analysed by HPLC, results may differ substantially from actual food ingestion if Chl a and carotenoids degrade at different rates during gut passage. As illustrated in the present study the proportions of the different pigments initially present in animals were similar to those of the food source, but the pigment concentrations decreased within minutes and at different rates affecting the proportions of the different pigments. Such a time and food-concentration-dependent change in pigment proportions will affect calculations of food selection using CHEMTAX. Therefore, a prerequisite for this approach will be very fast handling of animals upon collection to minimize the differential degradation of pigments, especially during diatom blooms. The present study clearly shows the importance of minimizing the handling time of the copepods (Fig. 3) from their experimental environment until they are frozen in total darkness. Results obtained this way should still be interpreted with caution due to factors like food-concentration-dependent recovery of pigments in the animals and ingestion-rate- and starvation-period-related degradation of pigments (Tirelli and Mayzaud, 1998).

The present study shows that pigments in principle can be used in a satisfactory way to detect food selection between phytoplankton types with different pigment compositions especially at higher food concentrations, but in order to be able to get a good signal very fast handling of samples is essential. Care has to be taken if the copepods contain astaxanthin or astaxanthin esters as those can interfere with some phytoplankton pigments, which makes these pigments not optimal to use for estimation of food selection. The present study also showed that faecal pellets represent the available food even better than gut content, again, at high food levels. Faecal pellets do not have the interference of body pigmentation, and could therefore potentially be a better indicator for selected food type, but this should be further investigated.

Our simple analysis of food condition indicates that use of pigments in estimating natural food selection might depend on food availability and zooplankton concentration, and is therefore site and season specific. As an example, a conservative food to predator ratio on the Dogger Bank, in the North Sea, during summer is, e.g. for Temora 3 ind. L−1 in 2.5 µg Chl a (Koski et al., 2011) that gives a Cfood:Czoo ratio of 3.2 and for Calanus 0.5 L−1 at the same food level (Jónasdóttir and Koski, 2011) results in a ratio of 1.9, both of which should give a satisfactory pigment signal. It should also be kept in mind that most often zooplankton have diurnal feeding rhythm, and therefore collection of zooplankton for gut content analysis should happen during, or at the end of the daily feeding cycle. Our gut evacuation experiment showed that the relative composition of the gut pigments remains fairly stable only the first 15–20 min after feeding, after which degradation of specific pigments will cause a skewed picture of the composition.

FUNDING

This study was funded by the Danish Research Agency grant number: 21-03-0487 to S.H.J. and P.H.

ACKNOWLEDGEMENTS

We thank J. Melbye (DTU-AQUA) for assistance with cultures of copepods and W. Martinsen and B.L. Møller (DCE) for help with HPLC analyses and L. Schlüter and anonymous reviewers for helpful comments on the manuscript.

REFERENCES

Besiktepe
S.
Dam
H. G.
Coupling of ingestion and defecation as a function of diet in the calanoid copepod Acartia tonsa
Mar. Ecol. Progr. Ser.
 , 
2002
, vol. 
229
 (pg. 
151
-
164
)
Chesson
J.
The estimation and analysis of preference and its relationship to foraging models
Ecology
 , 
1983
, vol. 
64
 (pg. 
1297
-
1304
)
Descy
J. -P.
Frost
T. -M.
Hurley
J. -P.
Phytoplankton biomass, production and grazing mortality in Exmouth Gulf, a shallow embayment on the arid, tropical coast of Western Australia
J. Plankton Res.
 , 
1999
, vol. 
21
 (pg. 
127
-
145
)
Frost
B. W.
Effects of size and concentration of food particles on the feeding behavior of the marine planktonic copepod Calanus pacificus
Limnol. Oceanogr.
 , 
1972
, vol. 
17
 (pg. 
805
-
815
)
Hansen
P. J.
The red tide dinoflagellate Alexandrium tamarense: effects on behaviour and growth of a tintinnid ciliate
Mar. Ecol. Progr. Ser.
 , 
1989
, vol. 
53
 (pg. 
105
-
116
)
Head
E. J. H.
Gut pigment accumulation and destruction by arctic copepods in vitro and in situ
Mar. Biol.
 , 
1992
, vol. 
112
 (pg. 
583
-
592
)
Head
E. J. H.
Harris
L. R.
Chlorophyll and carotenoid transformation and destruction by Calanus spp. grazing on diatoms
Mar. Ecol. Progr. Ser.
 , 
1992
, vol. 
86
 (pg. 
229
-
238
)
Head
E. J. H.
Harris
L. R.
Feeding selectivity by copepods grazing on natural mixtures of phytoplankton determined by HPLC analysis of pigments
Mar. Ecol. Progr. Ser.
 , 
1994
, vol. 
110
 (pg. 
75
-
83
)
Jeffrey
S. W.
Mantoura
R. F. C.
Wright
S. W.
Phytoplankton pigments in oceanography
 , 
1997
Paris
UNESCO Publishing
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.
 , 
2011
, vol. 
33
 (pg. 
85
-
103
)
Kleppel
G. S.
Frazel
D.
Pieper
R. E.
, et al.  . 
Natural diets of zooplankton off southern California
Mar. Ecol. Progr. Ser.
 , 
1988
, vol. 
4952
 (pg. 
231
-
241
)
Kleppel
G. S.
Willbanks
L.
Pieper
R. E.
Diel variation in body carotenoid content and feeding activity in marine zooplankton assemblages
J. Plankton Res.
 , 
1985
, vol. 
7
 (pg. 
569
-
580
)
Koski
M.
Jónasdóttir
S. H.
Bagøien
E.
Biological processes in the North Sea: vertical distribution and reproduction of neritic copepods in relation to environmental factors
J. Plankton Res.
 , 
2011
, vol. 
33
 (pg. 
63
-
84
)
Mackey
M. D.
Mackey
D. J.
Higgins
H. W.
, et al.  . 
CHEMTAX - A program for estimating class abundances from chemical markers: application to HPLC measurements of phytoplankton
Mar. Ecol. Progr. Ser.
 , 
1996
, vol. 
144
 (pg. 
265
-
283
)
Mauchline
J.
The Biology of Calanoid Copepods
 , 
1998
San Diego
Academic Press
Menden-Deuer
S.
Lessard
E. J.
Carbon to volume relationships for dinoflagellates, diatoms, and other protist plankton
Limnol. Oceanogr.
 , 
2000
, vol. 
45
 (pg. 
569
-
579
)
Meyer-Harms
B.
Irigoien
X.
Head
R.
, et al.  . 
Selective feeding on natural phytoplankton by Calanus finmarchicus before, during, and after the 1997 spring bloom in the Norwegian Sea
Limnol. Oceanogr.
 , 
1999
, vol. 
44
 (pg. 
154
-
165
)
Meyer-Harms
B.
von Bodungen
B.
Taxon-specific ingestion rates of natural phytoplankton by calanoid copepods in an estuarine environment (Pomeranian Bight, Baltic Sea) determined by cell counts and HPLC analyses of marker pigments
Mar. Ecol. Progr. Ser.
 , 
1997
, vol. 
153
 (pg. 
181
-
190
)
Pandolfini
E.
Thys
I.
Leporcq
B.
, et al.  . 
Grazing experiments with two freshwater zooplankters: fate of chlorophyll and carotenoid pigments
J. Plankton Res.
 , 
2000
, vol. 
22
 (pg. 
305
-
319
)
Penry
D. L.
Frost
B. W.
Chlorophyll a degradation by Calanus pacificus: dependence on ingestion rate and digestive acclimation to food resources
Limnol. Oceanogr.
 , 
1991
, vol. 
36
 (pg. 
147
-
159
)
Quiblier
C.
Bourdier
G.
Amblard
C.
, et al.  . 
Separation of phytoplanktonic pigments by HPLC for the study of phyto-zooplankton trophic relationships
Aquat. Sci.
 , 
1994
, vol. 
56
 (pg. 
29
-
39
)
Quiblier-Llobéras
C.
Bourdier
G.
Amblard
C.
, et al.  . 
A qualitative study of zooplankton grazing in an oligo-mesotrophic lake using phytoplanktonic pigments as organic markers
Limnol. Oceanogr.
 , 
1996
, vol. 
41
 (pg. 
1767
-
1779
)
Strathmann
R.
Estimating the organic carbon content of phytoplankton from cell volume or plasma volume
Limnol. Oceanogr.
 , 
1967
, vol. 
12
 (pg. 
411
-
418
)
Swalethorp
R.
Kjellerup
S.
Dünweber
M.
, et al.  . 
Grazing, egg production, and biochemical evidence of differences in the life strategies of Calanus finmarchicus, C. glacialis and C. hyperboreus in Disko Bay, western Greenland
Mar. Ecol. Prog. Ser.
 , 
2011
, vol. 
429
 (pg. 
125
-
144
)
Thys
I.
Leporcq
B.
Descy
J. P.
Seasonal shifts in phytoplankton ingestion by Daphnia galeata, assessed by analysis of marker pigments
J. Plankton Res.
 , 
2003
, vol. 
25
 (pg. 
1471
-
1484
)
Tirelli
V.
Mayzaud
P.
Gut pigment destruction by the copepod Acartia clausi
J. Plankton Res.
 , 
1998
, vol. 
20
 (pg. 
1953
-
1961
)
Tirelli
V.
Mayzaud
P.
Relationship between functional response and gut transit time in the calanoid copeod Acartia clausi: role of food quantity and quality
J. Plankton Res.
 , 
2005
, vol. 
27
 (pg. 
557
-
568
)
Wong
C. K.
Liu
X. J.
Siu
Y. Y.
, et al.  . 
Study of selective feeding in the marine cladoceran Penilia avirostris by HPLC pigment analysis
J. Exp. Mar. Biol. Ecol.
 , 
2006
, vol. 
331
 (pg. 
21
-
32
)
Wright
S.
Jeffrey
S. W.
Mantoura
R.
, et al.  . 
Improved HPLC method for the analysis of chlorophylls and carotenoids from marine phytoplankton
Mar. Ecol. Progr. Ser.
 , 
1991
, vol. 
77
 (pg. 
183
-
196
)

Author notes

Corresponding editor: Roger Harris