The astaxanthin content of copepods grown under PAR light significantly increased when the diet of T-Isochrysis plus enriched torula yeast was supplemented with astaxanthin or astaxanthin precursors (Figure 1; F_3,16 = 371.6, p < 0.001). The highest astaxanthin content was observed for the copepods feeding on the diet supplemented with Haematococcus (3.10 ng copepod⁻¹ or ca. 619 μg g⁻¹ dry weight), and the lowest astaxanthin content was observed for copepods feeding on the diet supplemented with free astaxanthin, Carophyll^® Pink (1.27 ng copepod⁻¹; Tukey’s B p > 0.05; Figure 1). Copepods feeding on the diet supplemented with Spirulina meal had the highest zeaxanthin content (F_3,16 = 9.6, p = 0.002). The β-carotene content of copepods feeding on the diet supplemented with either Haematococcus or Spirulina meals was significantly higher than the content of copepods feeding on either of the other two diets (F_3,16 = 10.6, p = 0.001; Tukey’s B p > 0.05). Fucoxanthin content of copepods was similar with all diets.

A. atopus had a carotenoid composition similar to other copepods [23,41,42,43,44]. Copepods increased their carotenoid content up to five-fold depending on the availability of carotenoids in the diet which shows that copepods are able to profit from carotenoid enriched diets [23,44]. The level of increase is on the same scale to that observed for Nitocra lacustris, a marine copepod (5.5 μg dry weight) that increased its astaxathin content from 1.0 ng copepod⁻¹ when feeding on the prasinophyte Tetraselmis suecica to 3.6 ng astaxanthin copepod⁻¹ when feeding a formulated feed containing lycopene, α-carotene, β-carotene, lutein, phytoene and phytofluene [44].

A. atopus incorporated more astaxanthin supplied mainly in the esterified form (Haematococcus meal) than when supplied in free form Carophyll^® Pink. Astaxanthin is deposited mainly as esters and not as free astaxanthin in many crustaceans, particularly during ontogenetic development [37,43]. Copepods may either digest, assimilate or incorporate preformed astaxanthin esters more efficiently than free astaxanthin [45,46], or have a preference for the stereoisomer 3S,3′S that is more abundant in Haematococcus meal (>99% 3S,3′S [47]) than in Carophyll^® Pink (18.75% 3S,3′S [46]). The importance of variations in carotenoid assimilation dependent on food concentration to shape the degree of astaxanthin accumulation cannot be dismissed [48]. Nevertheless, it is likely that the observed differences in copepod astaxanthin content reflect the importance of food quality, in terms of stereoisomers or chemical forms of carotenoids, for its digestibility and incorporation by copepods. The absence of astaxanthin in Spirulina meal and the large content of astaxanthin in copepods feeding on Spirulina points to bioconversion from precursor carotenoids by the copepods, possibly from β-carotene which is present in large amount in the diet [49]. It has been demonstrated that some crustaceans are able to bioconvert zeaxanthin into astaxanthin via β-doradexanthin [50], yet this carotenoid was not identified in our analysis. Therefore, this pathway is not, apparently, used by A. atopus to obtain astaxanthin from the abundant zeaxanthin in the Spirulina meal.

Zeaxanthin and β-carotene were reduced by ca. 50%, and fucoxanthin was detected below the quantification limit in 96-h starved copepods during the acute toxicity tests under PAR and long wavelength UV light relative to cultured copepods (data not shown). These pigments were detected below the quantification level in starving copepods exposed to short wavelength UV light for 96-h. The decrease in the astaxanthin content of copepods starved for 96 h and exposed to PAR and long wavelength UV light ranged from around 40% (minimum of 37% for the diet supplemented with Haematococcus meal) to 63% (diet supplemented with Carophyll^® Pink; Figure 2). When exposed to short wavelength UV light, starving copepods exhibited a severe astaxanthin decrease, ranging from 77% (diet supplemented with Haematococcus meal) to 92% (diet supplemented with Carophyll^® Pink). The astaxanthin content of 96-h starved copepods was significantly dependent on both the diet type used in the cultures (F_3,38 = 260.1, p < 0.001) and the type of light exposure (F_2,38 = 150.0, p < 0.001). A significant interaction between diet and light was also found (F_6,38 = 16.7, p < 0.001).

The reduction of β-carotene and zeaxanthin by ca. 50% in the 96-h starved copepods under PAR and long wavelength UV light relative to the start of the experiments may be either the result of its direct metabolic use, or its bioconversion to compensate for astaxanthin loss due to starvation. The presence of β-carotene in trace amounts and the great reduction of astaxanthin in copepods exposed to short UV light suggest the destruction of pigments during their utilization as protectants against photooxidation, as described for tide pool dwelling harpacticoid copepods in high radiation environments [41]. Oxidative degradation of carotenoids occurs in various biological systems, and the rate of oxidation decreases from β-carotene or zeaxanthin to astaxanthin [51].

Carotenoids possess a polyene chain, which is a long conjugated double bond system forming the backbone of the molecule. The polyene chain may be terminated by cyclic end groups that contain oxygen-bearing substitutes (e.g., astaxanthin), and is responsible for the antioxidant activities of the carotenoids, both by quenching singlet oxygen [52], and scavenging radicals to terminate chain reactions [53]. This mechanism of UVR protection offered by carotenoids agrees with that proposed by Sommaruga [54], who suggested that the mechanism by which carotenoids function is likely not direct photoprotection (via reflectance or absorbance of UVR) but rather indirect photoprotection by scavenging ROS formed by UVR. Ringelberg [36] noted that carotenoid compounds are conspicuously found within copepod fat globules, carapace epidermis, ovaries, and eggs, which are areas critical to survival, and highly prone to the damaging effects of UVR induced ROS. The proposed main function of astaxanthin in crustaceans, especially in the form of astaxanthin esters, is to generally improve the antioxidant protection of storage lipids, also in situations where photoprotection is not required [55].

The protection offered by carotenoids may be especially important against lipid peroxidation in cellular membranes, and polar carotenoids like astaxanthin are known to play a role in preserving membrane structure and reducing lipid hydroperoxide levels [56]. In membranes, polar carotenoids appear to span the membrane with their polar end groups extending toward the polar regions of the membrane bilayer, spanning the membranes in a parallel fashion, increasing the order parameter of the membrane bilayer [57]. This, in turn, restricts the permeation of the membrane to polar molecules and ions [58,59]. Astaxanthin has a molecular length similar to membrane bilayers, and also possesses a ketone group at the C4 and C4′ positions of the terminal rings, which may act to further stabilize astaxanthin's membrane interactions with respect to the polar terminal groups [57]. By spanning the entire width of the membrane as proposed by Woodall et al. [39], astaxanthin would enhance antioxidant activity by providing protection throughout the entire depth of the membrane, interfering with the propagation of free radicals in the hydrophobic core, and quenching radicals generated at the surface of membranes [57].

Fucoxanthin, zeaxanthin and β-carotene were detected below the quantification level in copepods exposed to copper at a concentration of 1 to 5 μM under all light types (data not shown). Under both PAR and long UV radiation, the effect of the diet was more significant on copepod astaxanthin content than the copper effect (Table 1, Figure 2). However, under short wavelength UV light, the opposite was observed (F_3,24 = 246.1, p < 0.001 for diet effects and F_1,24 = 432.7, p < 0.001 for copper effects; Table 1). The interaction between dietary and copper effects on the astaxanthin content of copepods was significant under all light regimes, and the strongest interaction was observed for copepods under short UV light (Table 1). The decrease in the astaxanthin content of copepods exposed to copper relative to control copepods under PAR light was lower for copepods fed on either diet supplemented with Haematococcus meal (39%) or Spirulina meal (20%) than for copepods feeding on the unsupplemented diet (41%) or the diet supplemented with Carophyll^® Pink (47%, Figure 2). Nevertheless, copepods feeding on the diets supplemented with either Haematococcus or Spirulina had a higher absolute astaxanthin decrease, which suggests an increased use of astaxanthin when it is more available in the body tissues. The same pattern was observed for copepods tested under long UV light. The decrease in astaxanthin content of copepods exposed to copper under short UV light was similar with all diets, ranging from 78% for the unsupplemented diet to 72% for the diet supplemented with Spirulina meal.

Copper (Cu) is an essential micronutrient for all living organisms, being involved in cellular respiration, free radical defense and cellular iron metabolism. Yet, at elevated levels, Cu is toxic to organisms and there is evidence that in vivo formation of reactive oxygen species is a mechanism of copper toxicity [60]. The astaxanthin content of copepods starved for 96 h under PAR or long UV light was more dependent on the diet offered than on the exposure to low copper concentrations (1–5 μmol). The higher dependence of the astaxanthin content on copper exposure than on the original diet under short UV light, points to the role of copper as a synergistic or additive factor to the natural environmental stressor UV light. In turn, short UV light interacts with copper to increase its destructive potential, thus increasing the influence of copper on copepod astaxanthin content to a higher level than that of the diet. The strong decrease of the astaxanthin content of copepods during both copper and short UV light exposure may have resulted from oxidative degradation that caused the disruption and breakdown of the polyene chromatophore [61]. This bleaching of pigments in crustaceans is often observed when animals are exposed to trace metals.

Mortality of Amphiascoides atopus ranged from 0–4% in control artificial sea water (ASW) under both PAR and long UV light, and from 0–8% under short wave UV light. Clear dose-response relationships were observed for survival after 96 h exposure, although the intensity of response under both PAR and long wavelength UV light was modulated by the type of diet used to culture the copepods (Figure 3). The culture diet had a smaller effect on the response of copepod survival to the exposure of copper under short wave UV light than under the other light types (Figure 3).

Estimates of the LC₅₀ for copper were higher when the copper exposure occurred under PAR or long UV light (5.26–13.72 μM or 334.24-872.02 μg L⁻¹; Figure 4) than when the tests were conducted under short wave UV light (2.34–3.98 μM or 148.71–253.50 μg L⁻¹; Figure 4 and Table 1). The LC₅₀ values and the corresponding upper and lower limits were higher when the copepods had been cultured under diets supplement with either Haematococcus meal or Spirulina meal, irrespective of the type of light exposure. Under both PAR and long wavelength UV light, the culture diet supplemented with Carophyll enabled an LC₅₀ of copper that was intermediate between the unsupplemented diet (Figure 4, T-Isochrysis) and diets supplement with either Haematococcus meal or Spirulina meal.

The protection offered by astaxanthin was relevant, in terms of survival, when A. atopus was exposed to the ROS generating copper toxicant. A. atopus mortality varied directly with Cu concentration, and the LC₅₀ 96 h for copper increased with the increase of copepod carotenoid content. The higher mortalities were observed when the copepods were exposed both to Cu and short UV light, and only the diets supplemented with Spirulina or Haematococcus were able to ascribe some degree of protection to the copepods. Bell et al. [62], using Carophyll pink as a source of astaxanthin to Atlantic salmon, observed that the carotenoid had antioxidant functions, offering protection against lipid peroxidation. The absence of protection against both copper and short UV exposure in the present experiment may simply be the result of inefficient assimilation and incorporation of artificial astaxanthin by copepods.

A. atopus is less sensitive to Cu than Tigriopus brevicornis and Tisbe battagliae and more sensitive to Cu than Tigriopus californicus, T. japonicus and Tisbe holothuriae (see Table 2). A. atopus is also more sensitive to cadmium (LC₅₀ 96 h, 4.13 ± 0.75 μMol Cd; [63]) than to copper (LC₅₀ 96 h, 4.42–6.23 μMol Cu), although the lower limit of the LC₅₀ of Cu overlaps with that of Cd. The lower sensitivity to copper relative to cadmium is the reverse to that observed for Tigriopus japonicus [64] and for Tisbe holothuriae [65]. Nevertheless, the latter harpacticoids were more resistant to the toxicity of both metals than A. atopus (see Table 2). The resistance to metals differs widely among harpacticoid species, different life stages and different geographically isolated populations, and some variations may also have been introduced by the quality of food used in some studies. The standardization of culture procedures and food for copepods used in toxicological tests [66,67] is an important step to obtain reliable results that allow comparisons among species and life stages.

The diet-dependent UV light resistance in high shore harpacticoid copepod Tigriopus brevicornis has been noted [41] and it is likely that the presence of astaxanthin gives an adaptive advantage to populations of astaxanthin rich species in copper-enriched environments where most intertidal seaweed and macroinvertebrate populations are eliminated [68]. Conversely, it has been observed that some harpacticoid species that are not acutely sensitive to most common pollutants, greatly increase their sensitivity after chronic exposure, especially in terms of population parameters [69]. Such observations, and the results presented here, point to the importance of copepod health and nutrition status during and after being subjected to toxicants.

Cultures of Amphiascoides atopus Lotufo and Fleeger were obtained from laboratory cultures kindly supplied by Dr. John W. Fleeger. The cultures were established in 1992 and have been continuously maintained since that date in sediment free, 1 L Erlenmeyer flasks at room temperature (23 °C [76]). Adult female copepods reach up to 5 μg dry weight [77]. Since the members of this genus typically live in beaches with coarse sediment and are not associated with muddy environments, all experiments were conducted free of sediment. For the present experiments, cultures were started with more than 50 males and 50 females grown in artificial sea water at 30‰ (ASW, Instant Ocean^® Sea Salt, USA), and fed every three days with T-Isochrysis paste (100 μL L⁻¹; Brine Shrimp Direct, Odgen UT, USA) plus enriched torula yeast diet (10 mg L⁻¹; Microfeast Plus^® L-10, Provesta Corp., OK, USA). T-Isochrysis paste is rich in poly-unsaturated fatty acids, especially docosahexaenoic acid (DHA [78]), and enriched torula yeast offers a diet rich in fatty acids, protein, minerals and B-vitamins (manufacturer specifications). T-Isochrysis paste has been individually used to successfully culture A. atopus in the laboratory and during toxicological studies [63], and enriched torula yeast has been used to mass culture the harpacticoid copepod Tisbe sp. [79]. After a week, the adults were removed by sieving the cultures through a 125-µm aperture screen and the nauplii were allowed to grow under photosynthetic active radiation (PAR) with a photoperiod of 12 h dark:12 h light and were fed one of the following test diets: (i) T-Isochysis = T-Isochrysis paste (100 μL L⁻¹) plus enriched torula yeast (10 mg L⁻¹); (ii) Carophyll = T-Isochrysis paste (100 μL L⁻¹) plus enriched torula yeast (10 mg L⁻¹) plus free astaxanthin (0.5 mg L⁻¹ Carophyll^® Pink, Hoffmann-La Roche, Basel, Swizerland); (iii) Haematococcus = T-Isochrysis paste (100 μL L⁻¹) plus Haematococcus meal (5 mg L⁻¹ NatuRose™, Cyanotech Corp., HI, USA); (iv) Spirulina = T-Isochrysis paste (100 μL L⁻¹) plus Spirulina meal (5 mg L⁻¹ Spirulina Pacifica^®, Cyanotech Corp., HI, USA). The food items used as a supplement of T-Isochrysis paste were chosen for their ability to offer a diet rich in astaxanthin (3,3′-dihydroxy-β,β′-carotene-4,4′-dione) or astaxanthin precursors. Carophyll^® Pink is nutritionally poor since the microbeads contain astaxanthin embedded in a matrix of gelatine and carbohydrate, enveloped by maize starch, according to manufacturer specifications. Both Haematococcus meal and Spirulina meal have additional nutritive value because of their amino acid and vitamin content [45,49]. The carotenoid pigments present in the supplements were extracted and analyzed by high performance liquid chromatography (HPLC) as described below for copepod total extracts. T-Isochrysis paste is a source of fucoxanthin and β-carotene, enriched torula yeast contains a small amount of astaxanthin, Carophyll^® Pink is a source of highly concentrated free astaxanthin (ca. 8% weight) containing approximately 75% of the natural all-E isomer [46], Haematococcus meal is mainly a source of astaxanthin in the mono (70%) and di-esterified (10%) form [45], and Spirulina meal is mainly a source of zeaxanthin and β-carotene (Table 3). Before adding the supplement diet to the food suspension, Carophyll^® Pink, Haematococcus and Spirulina meals were dispersed in ASW using a sonifier (Branson Sonifier 450, 3 mm diameter probe, output set on 4, duty cycle on 60%; Branson Ultrasonics, Danbury CT, USA). Previous pilot culture experiments have shown that the food concentrations offered to the copepods were in excess of the amount ingested for a period of 3 days, and all cultures were continuously reproducing.

Adults were harvested from the experimental cultures after 2–3 weeks by sieving the culture medium through a 125-µm aperture screen and the retained newly molted females (<48 h) were sorted under a stereo-dissection microscope. By selecting cohorts of a single development stage (i.e., newly molted females), we have reduced variations in copepod carotenoid content related to copepod ontogenic stage [37]. Females were placed in filtered (2 μm) ASW for 4 h to empty their gut, and ASW was changed twice at 1.5 h intervals. Groups of 50 females grown on all types of diets were placed in 100 mL plastic vials for 96-h acute toxicity tests (see below). Groups of 200–300 females from all cultures were placed over glass GF/C Whatman filters, subjected to a gentle vacuum suction to remove culture medium and rinsed with ASW filtered through a 0.2 mm pore size filter. The filters with the females were placed in extraction vials for carotenoid extraction and analysis (see below).

The filters containing the copepods were placed in extraction vials on ice and pigments were immediately extracted in 1 mL 100% HPLC grade acetone (Fisher Scientific) using a sonifier cell disruptor for 30 s (Branson Sonifier 450, 3 mm diameter probe, output set on 4, duty cycle on 80%; Branson Ultrasonics, Danbury CT, USA). The extracts were bubbled with nitrogen gas to remove oxygen, incubated overnight in the dark at 4 °C and then filtered using syringe filters (Sun International; diameter 13 mm; 0.2 mm pore size) to remove debris. The filtered extracts were dried under nitrogen gas and the pigments were re-suspended in 200 mL acetone and stored in the dark at −80 °C before enzymatic hydrolysis of astaxanthin esters or direct pigment analysis, which took place within 24 h.

Half of the extracts from copepod cultures under the four types of diet and all pigment extracts of copepods subjected to the 96-h acute toxicity tests were hydrolysed by an enzymatic procedure to yield free astaxanthin. The enzyme used was cholesterol esterase from Pseudomonas fluorescens (Sigma-Aldrich, MO, USA) and the procedure followed Jacobs et al. [80].

The pigments in whole and hydrolysed extracts were analyzed using a Hewlett Packard 1100 High Performance Liquid Chromatograph (HPLC) with a 100 µL loop autosampler and a quaternary solvent delivery system coupled to a diode array spectrophotometer. The diode array detector was set at 436 nm for detection of all carotenoid pigments, and at 478 nm for the detection and quantification of astaxanthin and astaxanthin esters. Separation of pigments was performed by reversed-phase liquid chromatography using a Adsorbosphere™ C18 column (5 µm particle size silica, 250 mm × 4.6 mm i.d., Alltech Associates Inc., USA) coupled to a guard column. Analytical separation of pigments was achieved by gradient delivery of three mobile phases adapted from Kraay et al. [81]. Peak analysis was performed using the Hewlett Packard ChemStation software. Pigments were identified by comparing their retention time and absorption spectra with those of authentic standards (Sigma-Aldrich Inc., USA), and pigment concentration was calculated relating its peak area in the recorded chromatogram with the corresponding area of calibrated pigment standards. Pigment concentration was expressed as micromoles per copepod (μM individual⁻¹).

Copper solutions were prepared with Cu(II) chloride (reagent grade) and 30 ± 0.5‰ ASW previously filtered through a 0.2 µm pore size filter at 7 nominal concentrations: 0 (control), 1, 3, 5, 10, 15 and 20 μM (0, 63.55, 190.64, 317.73, 635.46, 953.19, 1270.92 μg L⁻¹). Water samples were collected after one hour of preparation and at end of each experiment in order to analyze the actual toxicant concentrations in the water. Samples were collected without disturbing the precipitate that could be present at the end of the experiment. The samples were analyzed on an Inductively Coupled Plasma–Optical Emission Spectrometer (ICP-OES, Perkin Elmer 2000 DV, Perkin Elmer Inc., USA). From the actual toxicant concentrations in the water at the start and the end of each survival test, average exposure concentrations were calculated assuming an exponential decrease of the toxicant concentration over time. All vials used to conduct bioassays were acid cleaned prior to use.

The copepods were subjected to three light treatments: PAR light with a photoperiod of 12 h dark: 12 h light (treatment 1); PAR light regime with the addition of 4 h of ultraviolet (UV) radiation with a wavelength of 365 nm (long wavelength or UV-A) at the middle of the PAR light period (treatment 2); PAR light regime with the addition of 4 h of ultraviolet (UV) radiation with a wavelength of 254 nm (short wavelength) at the middle of the PAR light period.

The 50 test female copepods were placed in each test vial using a glass pipette (four replicates per concentration and light treatment). After 96-h exposure, copepods were removed with a pipette to a shallow glass dish and the 50 test copepods were enumerated as live or dead. Copepods immobilized were considered alive if they displayed body motion when touched with a probe or exhibited clear motion of the digestive tract.

A 50% lethal concentration (LC₅₀) and 95% confidence interval for the contaminant were calculated from dose response mortality data by a non-linear curve-fitting procedure using Probit analysis (SPSS Statistical Package, SPSS Inc., USA). All LC₅₀ estimates from sediment toxicity tests were based on measured contaminant concentrations.

After the 96-h acute toxicity tests, three or four groups of 80–100 live copepods pooled from the 5 μM nominal copper concentrations were individually picked from the counting vials with a glass pipette, placed on GF/C Whatman filters, subjected to a gentle vacuum suction to remove the medium and rinsed with ASW filtered through a 0.2 mm pore size filter. Pigment extraction followed the same procedure applied to copepods from cultures.

ANOVA were applied to astaxanthin molar data of both algae and copepods (SPSS Statistical Package, SPSS Inc., USA). Data were square root transformed to meet the requirement of homoscedasticity when group variances were proportional to the means. Unplanned comparisons were made using Tukey’s B test (α = 0.05) or Hochberg’s GT2 tests (in the case of unequal sample size) to identify homogeneous sub-sets.

Survival was expressed as percentage of the corresponding controls and plotted against the actual toxicant concentration in the seawater. From the obtained dose response plots, LC₅₀ values and their corresponding 95% confidence limits were calculated by a non-linear curve-fitting procedure using Probit analysis (SPSS Statistical Package, SPSS Inc., USA).