A short life-cycle test methodology is presented for assessing toxic effects to the survival, reproduction (as gravidity), and development of the epibenthic harpacticoid copepod species N. spinipes. In the first phase of the test (4-d survival-gravidity), gravidity for controls was consistently 70 to 90% in water-only tests and 30 to 80% in the sediment exposures. Although gravidity was less than desirable for some of the control sediments, for both waters and sediments determining significant differences in reproductive and developmental endpoints between controls and contaminated media was possible. The lower gravidity in the control sediments may be attributed in part to lower chances of males encountering females, although the iteroparous females had the ability to become gravid in the absence of males.
Sensitivity to contaminants
At dissolved Cu concentrations of 100 µg/L or higher, toxic effects to N. spinipes reproduction and development were consistently observed. Of the various endpoints assessed, the most sensitive stage of the life cycle was N/G (EC10 = 51 µg Cu/L), followed closely by C/G and gravidity, with EC10 values of 90 and 130 µg Cu/L, respectively (Table 2). Survival, with an EC10 of 256 µg Cu/L, was the least sensitive endpoint. The results indicated that N. spinipes is of comparable sensitivity to other species commonly used for whole-sediment toxicity test procedures 5, 27. Considering copepod species, Bengtsson 14 reported an LC50 value for N. spinipes survival of 1,800 µg Cu/L, which is five times greater than the value of 347 µg Cu/L obtained in the present study. For the earlier studies, the N. spinipes were isolated from sediments of a salinity of 7‰ and had presumably originated from brackish waters of the Baltic Sea, off Stockholm, Sweden. Differences in the copepod isolates, and the culturing and testing conditions, are likely causes of the differences between the results of Bengtsson 14 and those of the present study. Considering other harpacticoid species that have been used for sediment toxicity tests, Tigriopus japonicus exhibits a 96-h LC50 of 2,200 µg Cu/L 28, and A. tenuiremis exhibits a 96-h LC50 value of 124 µg/L 9. The most sensitive harpacticoid species reported is Tisbe battagliai (96-h LC50 of 88 µg/L, 29); however, sediment tests have not been developed using this species. Ward et al. 15 noted that a similar harpacticoid, Tisbe tenuimana, was also very sensitive to Cu (24-h LC50 of ∼50 µg/L), but the species was not considered useful for sediment tests because of its small size and difficulties in recovery from sediments.
Comparison of endpoints
In the short life-cycle test, the first stage (4-d adult survival/%gravid females) showed a clear concentration–response relationship; however, no such clear response for the percentage of gravid females results (Fig. 2a) was seen, particularly in the 0 to 200 µg/L Cu concentration range. Other experiments presented in Supplemental Data, Fig. S1a suggest that the variability in the percentage of females becoming gravid was not likely to be attributable to the Cu exposure, because most of these experiments achieved gravidity values similar to or higher than controls at 50 µg Cu/L. These results indicated that 4-d survival, although less sensitive, was more robust than 4-d gravidity. Likely gravidity was being affected by some underdetermined confounding factors.
In the second stage (7-d development test, Fig. 1), the egg sacs of the preexposed gravid females were produced through mating that occurred while the male and female couples were exposed to Cu, and the nauplii immediately exposed to Cu from hatching. In contrast, when non-preexposed gravid females were used to initiate the 7-d development phase of the test, the contaminant exposure occurred after egg production, commencing partway through the egg-to-hatching development stage. Despite these differences in Cu exposure, for the lower Cu concentrations (e.g., 50 and 100 µg Cu/L), a greater number of offspring was produced during the 7-d development phase from preexposed gravid females than from non-preexposed gravid females. This also indicated that factors other than the Cu exposure contributed to the lower gravidity. These differences could be related, in part, to biological variability between the preexposed gravid females and the non-preexposed gravid females used in the different tests, because the control values were different as well. Therefore, the differences between preexposed and nonexposed gravid treatments should only be compared in terms of EC50 values and not of net number of nauplii or copepodites per gravid female. Although the EC50s for N/G and total offspring per gravid female were not significantly different for the two exposures (Table 2), the EC50s for C/G were significantly smaller for the preexposed gravid females (Table 2).
These results suggested that the early exposure of eggs to Cu may indeed influence the effects observed. Surprisingly, this Cu exposure appeared to stimulate the development of egg to nauplius for the low range of Cu concentrations (1–100 µg Cu/L). However, the preexposure to Cu caused negative effects when the concentrations were greater than 100 µg Cu/L. Although the presence of multiple broods during the 7-d development phase made interpretation of some results difficult, the more sensitive period of the life cycle appeared to be the nauplius-to-copepodite stage (estimated by the C/G endpoint). However, the sensitivity of this stage appeared to be critically dependent on the full reproduction–development stage being subjected to the exposure.
The results of the multiple-generation test indicated that a full life-cycle exposure to Cu does not increase the sensitivity of the next generation. The sensitivities of both development endpoints (N/G and C/G) were higher for the first generation (9-d development test phase) than for the second generation. Possibly the small differences in diet for the first generation (which had a diet in cultures of mixed algae) and second generation (which had a diet during the tests of fish food) may have affected the egg-to-nauplius development and nauplius-to-copepodite development stages differently. These results suggested that the exposure of a whole generation to up to 100 µg Cu/L has negligible effects in the survivorship of the next generation. This result was consistent with the observations from similar multiple-generation tests with the T. japonicus19 and for T. angulatus, where the intrinsic rate of growth during a 28-d Cu exposure was observed to be not statistically different from that of controls at 20 to 60 µg/L, but was significantly lower at concentrations of 103 and 180 µg/L 30.
Because of the low number of copepodites produced at the highest Cu concentrations, the reproduction phase of the multiple-generation test (from new copepodite to gravid females) exhibited less conclusive results. At less than 100 µg Cu/L, the survival remained high and was not significantly different from control, but gravidity was greater than in the controls. This result, also observed in the short life-cycle test, suggested than low doses of Cu might stimulate the production of broods and also the development of egg to nauplius and copepodite stages. Such stimulation may be explained by hormesis, in which overcompensatory biological responses occur as a response to toxicants.
Hormesis has been defined as a biphasic dose–response phenomenon characterized by a low-dose stimulation and a high-dose inhibition 31–33. This phenomenon has been widely reported in studies for numerous species and endpoints 33. Hormetic effects also have been reported previously for toxicological studies with copepods 7, 34. Based on the analyses of thousands of dose–response relationships with evidence of hormesis, Calabrese and Baldwin 33 concluded that the maximum likely stimulatory response was 30 to 60% greater than controls. However, N. spinipes exhibited much greater stimulatory effects in the current study. For the multiple-generation tests, the reproduction at low Cu doses was frequently stimulated to more than 60% greater than controls (see Supplemental Information, Fig. S1). Furthermore, the stimulatory response was even higher in the development endpoints, with development being 250 to 450% and 270 to 370% higher than in controls for N/G and C/G, respectively.
The 39-d multiple-generation exposure was expected to provide a more realistic approach to assess life-cycle effects in copepods. However, the results obtained in this study have shown that long-term tests can be strongly affected by confounding factors that greatly reduce the benefits of a longer test duration. Many possible confounding factors have been previously summarized in the literature 35; high nauplii mortality in controls being the most common 12.
Overall, the short life-cycle test proposed incorporates a suite of advantages in comparison with full life-cycle tests. The controls achieved greater than 80% survival and 70 to 90% gravidity (for waters), and the test endpoints were of similar sensitivity to Cu to those of the full life-cycle test (≥ 21 d) in which the gravidity was typically only 30 to 60% (Figs. 2 and 3). These results are also consistent with those of other studies; for example, 21-d bioassays were not more sensitive than 14-d bioassays for the copepods Microarthridon littorale and A. tenuiremis35.
A significant advantage of the short life-cycle test was that it was less affected by confounding factors such us feeding quality or hormetic stimulation in the low-range concentrations. Recently, researchers observed for N. spinipes that some microalgae commonly used as food, in particular, Dunaliella tertiolecta and Phaeodactylum tricornutum, may result in low reproduction, poor juvenile survival, and compromise nauplii development, when compared with feeding with the algae Rhodomonas salina18. The food type and quantity used in the current study for all of the toxicity tests were based on a previous study 15, which showed that for a short exposure time (7 d) the use of fish food (as used for all tests in the present study) as nutrition produced similar endpoint values as a tri-algal diet (as used for culturing N. spinipes in the current study). However, although the feeding regimen of powdered fish food had been chosen as optimal based on a 7-d gravidity test, this may not have been optimal for the longer experiments.
The results indicated that the short N. spinipes life-cycle test was amenable to whole-sediment toxicity tests. Although survival was greater than 80%, the gravidity in sediment controls was typically 30 to 80%, compared with 70 to 90% for the equivalent water-only test. Gravidity in the sediment tests was, however, similar to that achieved in the multiple-generation test water–only controls.
Comparing LC50 values for sediment exposures with those from other studies is more difficult, because of differences in the partitioning of contaminants between water and sediment phases and also differences in organism exposure routes 36. In the present study, the LC50 for survival was 1,900 mg Cu/kg for Cu-spiked sediments. This was considerably higher than the LC50 of 281 mg Cu/kg reported for the copepod Amphiascus tenuiremis9 or the LC50s of 60 and 534 mg Cu/kg reported for the amphipods Ampelisca abdita and Eohaustorius estuaries, respectively 5. For A. tenuiremis, the dissolved Cu exposure was likely to have contributed to the observed effects and an LC50 of 125 µg Cu/L was calculated based on porewater Cu concentration 9. In the amphipod tests, dissolved Cu concentrations in the overlying water were 430 to 2,920 and 5,300 to 13,900 µg/L for A. abdita and E. estuarius, respectively. Because the LC50 value for A. abdita was approximately 20 µg Cu/L, the toxic effects were more likely to be from the dissolved Cu than from the particulate Cu. In the current study, water overlying the Cu-spiked sediments contained less than 100 µg Cu/L, and toxic effects were likely to also have been primarily through dietary exposure to particulates 37.
The degree of contamination of naturally contaminated sediments would be considered as moderate to high for I and III and very high for II and IV 1, 38. Although survival was reduced to below 60% in only the sediments with very high contamination, gravidity was significantly reduced (< 20% of controls) in all of the contaminated sediments tested. The development endpoints were significantly affected, especially for sediments II, III, and IV, where no offspring were observed.
The results indicate that the application of the short life-cycle test to sediments will provide useful information on chronic effects of sediment contaminants on organism life-cycles. The observation of no offspring survival in some contaminated sediments may indicate that N. spinipes may be more susceptible to effects in whole-sediment tests than in water-only tests. The sensitivity of the developmental endpoint may be higher in the presence of sediment, through lower gravidity or additional stress for the released nauplii. However, one might expect that harpacticoid copepods should be quite tolerant of physical stress caused by sediments. An alternative explanation is that exposure to contaminants through diet may be very important in the sediment tests, a route that does not exist in the water-only tests that only receive dietary exposure to contaminants that adsorb to the added food. Similar observations have been made for whole-sediment tests using the epibenthic deposit feeding amphipod, Melita plumulosa21.