Comparison of Chironomus riparius Meigen and Chironomus tentans Fabricius (Diptera: Chironomidae) for assessing the toxicity of sediments
Abstract
The benthic macroinvertebrates Chironomus riparius Meigen and C. tentans Fabricius were compared to evaluate their role in the assessment of sediment toxicity. Larval growth and adult emergence of both species were used to assess the toxicity of two sediments, one natural and one artificial, spiked with either cadmium or lindane (i.e., γ-hexachlorocyclohexane). Both toxicants significantly reduced (p < 0.05) larval growth and survival in relation to control animals, with C. tentans being the more sensitive species. Cadmium spiking had no effect on the number of adults that emerged for either species, but emergence times were delayed at the highest cadmium concentration. In contrast, lindane spiking produced a significant reduction (p < 0.05) in adult numbers in relation to control animals, whereas emergence times were not affected. In both the growth and emergence tests, toxicity differed depending on the choice of species and the sediment type. In terms of comparative sensitivity, C. tentans was not only more susceptible to both toxicants but seemed to be less physically robust than C. riparius, leading to some variable data, especially in the emergence study.
INTRODUCTION
Both larval growth and adult emergence of Chironomus riparius and C. tentans have been used as sublethal, chronic response criteria to investigate sediment toxicity, and several authors have reported a reduction in one or both of these parameters in response to sediment contamination [1-4]. The major attribute of these two test endpoints, which has contributed to their widespread use, is sensitivity. This was highlighted by Giesy et al. [5] in a comparative test involving three river sediments in which growth retardation of larval C. tentans was more sensitive than Microtox® (Microsoft, Redmond, WA, USA) and Daphnia sp. assays. This sensitivity may reflect the fact that successful larval growth and eventual adult emergence require the integration of biochemical, physiological, and behavioral processes, which are known to be affected by the environment [6] and by pollutant stress [7]. Another contributory factor is the extended exposure period, with growth typically being assessed over the course of 10 d and emergence from 15 to 30 d being dependent on the test species (i.e., C. riparius or C. tentans). Thus, most (or at least a large proportion) of the life cycle, including the sensitive first-instar larvae in the case of emergence tests, is incorporated in the study.
The suitability of these endpoints in the study of sediment-related problems using C. riparius and C. tentans is supported through their recommendation for use by various testing agencies in Europe, North America, and Canada [8-10]. Recently, these endpoints have found increasing use in dose-effect sediment tests, largely because of the need to obtain a better understanding regarding the complex interactions within the sediment-water system in relation to contaminant binding, accumulation, routes of organism exposure, and in particular, bioavailability. Such information is essential to environmental regulators and policymakers alike so that the threat posed to the aquatic environment by polluted sediments can be assessed accurately and any improvements monitored.
The use of either species is acceptable, but C. riparius is preferentially used among European agencies and C. tentans among North American agencies. This can lead to problems when interpreting results generated from tests conducted using separate species, because little direct comparative data exist. Burton et al. [11], however, identified C. riparius as being more sensitive than C. tentans after exposure to several contaminated sediments. Problems of comparison are compounded by the selection of the life-cycle stage to be used in these tests. Some consistency does apply with first-instar larvae generally being used to initiate emergence studies [12, 13], but this does not occur with larval growth tests. Several groups [2, 14] have demonstrated the reliability and sensitivity of using second-instar C. riparius larvae, whereas others [9, 10] have stipulated the use of third-instar C. tentans or first-instar C. riparius larvae. Because larvae show increasing tolerance as they develop, it seems logical that for direct comparison, the same life stage should be used regardless of species, thereby eliminating any potential confounding effects that may arise because of differing sensitivities.
Considering these issues, the current investigation aimed to compare two chronic response criteria of C. riparius and C. tentans as a means of assessing the toxicity of sediments. To achieve this, one natural and one artificial sediment were spiked at three separate concentrations with either cadmium or lindane, and their effects on both larval growth and adult emergence of the test species were recorded. The data are analyzed to compare species response to toxicant concentration and sediment type, and differences were evaluated in terms of the possible implications for sediment assessment.
| Nominal concentration | Measured concentration |
|---|---|
| 0.6 mg cadmium/La | 0.59 mg/L |
| 1.25 mg cadmium/La | 1.18 mg/L |
| 2.5 mg cadmium/L | 2.41 mg/L |
| 5 mg cadmium/L | 4.84 mg/L |
| 10 mg cadmium/L | 9.66 mg/L |
| 35 μg lindane/L | 16.6 μg/L |
| 75 μg lindane/L | 31.6 μg/L |
| 150 μg lindane/L | 87.7 μg/L |
- a These concentrations were only used in the emergence test.
MATERIALS AND METHODS
Test sediments
One natural and one artificial sediment were used in both the growth and the emergence studies. The natural sediment, a terrestrial soil supplied by Zeneca Agrochemicals (Jealotts Hill, Bracknell, UK), was composed of 12% sand, 31% clay, and 57% silt. This sediment was dried and sieved to 2 mm before use. The artificial sediment was based on a combination of recommendations from the Organization for Economic Cooperation and Development [15] as well as Suedel and Rodgers [16]. Essentially, the artificial sediment was composed of 15% organic matter (Shamrock®, Bord na Móna, Newbridge, Ireland; peatfree wood-chip compost dried at 100°C with large debris removed), 80% sand (horticultural-grade silver sand subdivided into coarse [2.0–0.5 mm], medium [0.5–0.25 mm], and fine [≤0.25 mm] fractions), and 5% kaolin clay.
Sediment spiking
The two test sediments were spiked for 4 h with three concentrations of cadmium or lindane using the suspension technique [17, 18]. The same spiking concentrations of lindane (16.6, 31.6, and 87.7 μg/L) were used in both the growth and emergence tests. However, after the cadmium growth test, in which spiking concentrations of 2.41, 4.84, and 9.66 mg/L were used, a lower range of concentrations (0.59, 1.18, 2.41 mg/L) was deemed to be necessary for the emergence test. Analysis of the sediment spiking solutions confirmed a good agreement between the nominal and actual concentrations (Table 1).
Sediment chemistry
For the growth and emergence studies, samples of overlying water, pore water, and sediment were collected from blank replicates (water and sediment but no larvae) to avoid disrupting the test replicates. Samples were taken at the start, the midpoint, and the end of the respective experiments. Since concentrations did not deviate with respect to time or the type of test, only a representative sample of the data is provided in Table 2. In the cadmium-spiked sediments, most of the metal was associated with the sediment phase. In the lindane-spiked sediments, this was probably also the case, but because of the poor performance of the extraction technique, this was not evident from the data.
Overlying water. Samples were taken from above the sediment and, in the case of cadmium, fixed at 1% with Aristar® nitric acid (BDH, Poole, UK). Lindane samples were extracted into n-hexane by shaking for 1 to 2 min before analysis. All samples taken on day 1 after spiking had to be filtered through a 0.45-μm filter to remove particulate material, but this was not necessary for the subsequent samples.
| Measured toxicant concentration | |||
|---|---|---|---|
| Sediment treatment | Overlying water (mg cadmium/L) | Pore water (mg cadmium/L) | Sediment (mg cadmium/L) |
| Artificial | |||
| 0.59 mg/L | 0.006 | 0.014 | 0.106 |
| 1.18 mg/L | 0.008 | 0.055 | 0.106 |
| 2.41 mg/L | 0.002 | 0.076 | 0.39 |
| 4.84 mg/L | 0.003 | 0.274 | 1.33 |
| 9.66 mg/L | 0.006 | 0.477 | 2.07 |
| Natural | |||
| 0.59 mg/L | 0.002 | 0.004 | 0.388 |
| 1.18 mg/L | 0.003 | 0.007 | 0.596 |
| 2.41 mg/L | 0.004 | 0.030 | 1.18 |
| 4.84 mg/L | 0.008 | 0.031 | 2.62 |
| 9.66 mg/L | 0.017 | −a | 5.32 |
| Overlying water (μg lindane/L) | Pore water (μg lindane/L) | Sediment (μg lindane/g) | |
| Artificial | |||
| 16.6 μg/L | 0.08 | 0.49 | 1.38 |
| 31.6 μg/L | 0.23 | 1.20 | 1.71 |
| 87.7 μg/L | 0.25 | 2.25 | 2.07 |
| Natural | |||
| 16.6 μg/L | 0.37 | 0.49 | 2.07 |
| 31.6 μg/L | 0.87 | 0.63 | 2.40 |
| 87.7 μg/L | 1.89 | 1.69 | 3.42 |
- a Sample lost before analysis.
Pore water. Pore water was extracted by centrifugation of the wet sediments at 3500 rpm for 40 min at 4°C in a Sorvall® RC-5B refrigerated super-speed centrifuge (Sorvall Products, Neartown, CT, USA).
Sediment. Sediment samples were heated at 100°C or less to constant weight, and 1 g of dried sediment was then used in the analysis. Cadmium-spiked sediments were transferred to Pyrex® (Corning Corporation, Corning, NY, USA) tubes containing 4 ml of Aristar nitric acid and then placed into a heating block at 70°C for a maximum of 4 h. Tubes in which digestion was completed (as indicated by a clear yellow solution) before this maximum time elapsed were removed. Lindane was extracted from sediment by manual shaking for 2 min with 3 ml of n-hexane in a 20-ml Quickfit® (Bibby Sterilin, Staffordshire, UK) glass tube. The extract was allowed to settle overnight, transferred to a separate tube, and then subjected to further shaking (1 min) before analysis.
Toxicant analysis
Cadmium concentrations were determined by atomic absorption spectrophotometry against suitable standards (0.1–2.0 mg/L) on an Instrumentation Model 457 (Instrumentation Laboratory, Lexington, MA, USA) using standard operating procedures. Lindane concentrations were analyzed by measurement of samples against a 10-μg/L standard by gas–liquid chromatography on a Pye Unicam 4500 (Unicam Scientific, Pt. Pleasant, FL, USA) with a 5% SE-30 packed column and an electron-capture detector.
Water quality
The dilution water (i.e., growth test) and the overlying water in the control vessels (i.e., emergence test) at the start, midpoint, and end of the respective tests were measured for the following water-quality parameters: dissolved oxygen, 89–82% of air saturation value; conductivity, 256–202 μS/cm; pH, 7.8–6.9; and hardness, 116–140 as mg CaCO3/L.
Growth test
Experimental design. Fifteen larvae each of C. riparius and of C. tentans were exposed at each concentration in individual soda glass vials (5 × 2 cm) at 22°C ± 1°C with a 16-h light photoperiod and an 8-h dark period. Each vial received 4 g (wet wt) of spiked sediment or non-spiked control before the addition of 10 ml of overlying dechlorinated water. To minimize disruption of the sediment, a polystyrene plate was placed over the sediment surface as the water was added.
After a 24-h settling period, 8 ml of the water column were removed and replaced with fresh dechlorinated water, a process that was repeated daily for the duration of the test to maintain the dissolved oxygen at 80% or greater of the air saturation value [19]. Individual second instar larvae of both species (5 d after hatch) were randomly selected and transferred to a vial. On addition, larvae were fed with 500 μl of a 1-g/L suspension of ground Tetramin® (TetraWerke, Melle, Germany) fish flake. Food was not subsequently provided for 48 h; thereafter, larvae were fed on a daily basis. This 48-h delay was employed because of previous work in our laboratory, which showed that during this stage of the test, relatively little food is consumed, presumably because of larval settlement and tube construction. At the end of the 10-d exposure period, larvae were carefully extracted from the sediment, and the individual wet weights were recorded using a sensitive, nondestructive technique [20, 21]. Larvae were then dried at 100°C or less overnight to a constant weight, and when possible, individual dry weights were recorded.
Data analysis. Data from the growth tests were analyzed using one-way analysis of variance to determine differences in final larval weights (dry and wet) and also underwent multiple comparisons (Tukey-Kramer) to locate differences between sediments. The nonparametric Kruskal-Wallis test, which was followed by a Tukey-type multiple comparison, was performed if the analysis by parametric methods was not suitable. Both wet and dry weights were reported, because the small size of some larvae (second instar) recovered from the test made accurate determination of individual dry weights almost impossible. Wet weight provides a sensitive indication of the effects on larval growth [14], but results may be confounded by larval water content. Reporting dry weights eliminates this problem and also facilitates comparison with data from other sediment tests in which dry weight is reported as a standard practice.
Emergence test
Experimental design. Larvae of C. riparius and C. tentans were exposed to sediment in test vessels identical to those described by Watts and Pascoe [13] at 22°C ± 1°C with a 16-h light photoperiod and an 8-h dark period. Three replicate vessels were used at each concentration for both species. Each vessel received 57 g (wet wt) of spiked natural or artificial sediment, or non-spiked controls of both, and 400 ml of overlying dechlorinated water. A polystyrene plate was positioned over the sediment to minimize dispersal of the sediment as the water was added. Gentle aeration of the water column was provided and maintained throughout the test, negating the need for solution changes. After a 48-h settling period, 20 first-instar larvae of C. riparius or C. tentans were randomly selected and assigned to each test vessel using a glass pipette. Individual larvae were checked for viability (i.e., swimming motion) in the water column as they were added. On addition, larvae were fed with a suspension of ground Tetramin fish flake (0.5 mg flake/larva). Thereafter, larvae were fed only when most of the previous ration had been consumed, which was necessary to avoid excess build up of food and an associated deterioration in water quality that had been noted during preliminary studies. The red/white coloration of the Tetramin provided a contrast to the darker sediment, thereby allowing a crude assessment of food consumption to be made.
Emergence of adults. Both the number and the sex of emerged adults were recorded daily before their removal from the test vessel. This practice continued until 100% emergence was achieved in a particular replicate or a period of 3 d had elapsed during which no emergence was recorded. Only those adults that had successfully broken free from the pupal skin were considered to have successfully emerged.
Data analysis. Adult emergence from each test system was analyzed using a FORTRAN program written in this laboratory and based on the time-response method of Litchfield [22] to calculate and compare median emergence times (EmT50). Statistical differences in adult numbers between treatments were analyzed using one-way analysis of variance.
RESULTS
Growth test
Both cadmium- and lindane-spiked sediments significantly reduced the growth of C. riparius and C. tentans larvae in a dose-related manner (Tables 3 and 4). Both wet and dry weights are reported in the tables, along with percentage larval survival and percentage reduction in weight compared with controls for each sediment type. This latter criterion is only presented as wet weight because of the lack of dry weight data in the cadmium study; thus, to maintain consistency, all statistical comparisons referred to in this section are based on final larval wet weights. All control larvae survived and were recovered at the end of the test, but those of C. riparius were significantly heavier (p < 0.05) than those of C. tentans. In addition, C. riparius larvae had developed to the fourth instar by the end of the 10-d study, whereas C. tentans larvae had only reached the third instar.
Cadmium. The greatest level of growth inhibition and larval mortality was seen in the cadmium-spiked sediments (Table 3). Except for C. riparius exposed at the lowest treatment concentration (2.41 mg/L) in the artificial sediment, no cadmium-exposed larva of either species developed past the second instar, indicating a major inhibition of development. In artificial sediment, growth of C. riparius larvae was significantly reduced (Kruskal-Wallis [KW], p ≤ 0.001), with subsequent multiple-comparison analysis (p = 0.05) revealing that whereas the growth of larvae exposed to cadmium was reduced at all concentrations, those exposed at 2.41 mg/L were significantly heavier than those exposed at 4.8 or 9.6 mg/L. In natural sediment, no significant difference in weight was recorded between any of the cadmium-exposed larvae, but the growth of these larvae was significantly reduced (KW, p ≤ 0.001) compared with control animals. Comparison of C. riparius growth in both sediment types revealed no significant difference (p ≥ 0.05) in larval weight between the respective controls or the two highest cadmium concentrations.
A similar result was obtained after analysis of C. tentans larval growth in artificial sediment, in which control larvae were significantly heavier (KW, p ≤ 0.001) than those in each of the three treatments. However, in natural sediment, comparison was only possible between the control larvae and those in the 2.41-mg/L treatment, in which growth was significantly reduced (KW, p ≤ 0.001). No larvae were recovered from the 4.8- or 9.6-mg/L treatments, and because no evidence of tube construction was noted, all larvae were assumed to have died.
| Percent survival | Mean larval wet weight (mg [SE])a | Mean larval dry weight (mg [SE]) | Percentage of control wet weight | |||||
|---|---|---|---|---|---|---|---|---|
| Treatment | C. riparius | C. tentans | C. riparius | C. tentans | C. riparius | C. tentans | C. riparius | C. tentans |
| Artificial | ||||||||
| Control | 100 | 100 | 6.64 (0.29) | 2.88 (0.29) | 1.13 (0.04) | 0.56 (0.05) | − | − |
| 2.4 mg/L | 100 | 80 | 4.07 (0.36) | 0.29 (0.05) | 0.72 (0.06) | 0.14 (0.04) | 61.3 | 10.1 |
| 4.8 mg/L | 100 | 73.3 | 0.29 (0.03) | 0.13 (0.01) | − b | − b | 4.4 | 4.5 |
| 9.7 mg/L | 86.7 | 80 | 0.27 (0.03) | 0.12 (0.01) | − b | − b | 4.1 | 4.2 |
| Natural | ||||||||
| Control | 100 | 100 | 7.10 (0.36) | 3.96 (0.35) | 1.09 (0.09) | 0.46 (0.03) | − | − |
| 2.4 mg/L | 100 | 66 | 0.64 (0.13) | 0.18 (0.01) | 0.13 (0.01) | − b | 9.0 | 4.5 |
| 4.8 mg/L | 93.3 | 0 | 0.44 (0.06) | − c | − b | − b | 6.2 | − c |
| 9.7 mg/L | 86.7 | 0 | 0.33 (0.03) | − c | − b | − b | 4.6 | − c |
- a SE = standard error.
- b Dry weight calculation was not possible because of the small size of the larvae.
- c No larvae were recovered (assumed 0% survival).
Lindane. In general, survival and final larval weights for both species (Table 4) were greater in the lindane than in the cadmium study, with virtually all C. riparius larvae, regardless of treatment type, reaching the fourth-instar stage. Most of the recovered C. tentans had reached the third-instar stage, but the frequency of the second-instar stage did increase with increasing lindane concentration. Larvae of C. riparius were significantly heavier (p ≤ 0.001) than larvae of C. tentans in all cases.
For C. riparius larvae exposed in artificial sediment, a significant difference (p ≤ 0.001) in the final larval weight was noted, with subsequent multiple-comparison analysis (p = 0.05) confirming that mean larval weight was significantly reduced only at the highest lindane concentration (87.7 μg/L). In natural sediment, control larval wet weight was significantly greater (p ≤ 0.001) than that at each of the three lindane concentrations. No significant difference in weight was noted between larvae exposed at 31.6 and 87.7 μg/L, but growth in the latter treatment was significantly reduced compared with that in the lowest exposure concentration (16.6 μg/L). Comparison of C. riparius growth in both sediment types revealed no significant difference (p ≥ 0.05) in wet weight between the respective controls or between the lowest and the highest lindane concentrations in the two sediments.
Analysis of C. tentans recovered from the artificial sediment revealed a significant difference (P ≤ 0.001) in the final larval weight, with multiple-comparison analysis (p = 0.05) confirming that control larvae were heavier than larvae in the remaining treatments. No significant difference (p ≥ 0.05) in weight was found between larvae exposed at any of the lindane concentrations. In natural sediment, C. tentans larval weights were significantly affected (KW, P ≤ 0.001), with nonparametric Tukey-type comparisons showing that control larvae were heavier in all cases. Of the lindane-exposed larvae, wet weights at the highest and the lowest concentrations were significantly different (p < 0.05), but neither differed significantly from that at the intermediate concentration. Comparison of larval wet weights between sediments revealed no significant difference (p ≥ 0.05) between any of the respective treatment groups.
Emergence test
In all cases, adults of C. riparius emerged significantly earlier (p < 0.05) than those of C. tentans, and for both species, a protandrous emergence pattern was noted. Analysis of the EmT50 (Table 5) confirmed this, with male adults of both species reaching their median emergence time significantly earlier than female adults in the same treatment group. Median emergence times are only provided for the control and the highest treatment concentration in both the artificial and natural sediments, because no difference in EmT50 (compared with controls) was noted at the intermediate concentrations. Cadmium exposure concentrations ranged from 0.59 to 2.41 mg/L. For the lindane-spiked natural sediment, the intermediate (31.6 μg/L) concentration is shown, because high mortality at the highest exposure concentration (87.7 μg/L) did not permit calculation of the EmT50.
| Percent survival | Mean larval wet weight (mg [SE])a | Mean larval dry weight (mg [SE]) | Percentage of control wet weight | |||||
|---|---|---|---|---|---|---|---|---|
| Treatment | C. riparius | C. tentans | C. riparius | C. tentans | C. riparius | C. tentans | C. riparius | C. tentans |
| Artifical | ||||||||
| Control | 100 | 100 | 7.21 (0.23) | 3.61 (0.28) | 1.26 (0.05) | 0.57 (0.04) | − | − |
| 16.6 μg/L | 100 | 100 | 6.43 (0.21) | 1.72 (0.33) | 1.07 (0.04) | 0.27 (0.05) | 89.1 | 47.6 |
| 31.6 μg/L | 100 | 100 | 6.73 (0.24) | 1.89 (0.22) | 0.99 (0.04) | 0.28 (0.02) | 93.3 | 52.4 |
| 87.7 μg/L | 100 | 73.3 | 5.37 (0.15) | 0.99 (0.18) | 0.73 (0.04) | 0.17 (0.04) | 74.5 | 27.4 |
| Natural | ||||||||
| Control | 100 | 100 | 7.89 (0.32) | 3.90 (0.34) | 1.44 (0.08) | 0.64 (0.07) | − | − |
| 16.6 μg/L | 100 | 86.7 | 6.21 (0.19) | 2.02 (0.26) | 1.12 (0.04) | 0.32 (0.05) | 78.7 | 51.8 |
| 31.6 μg/L | 100 | 80 | 5.21 (0.33) | 1.39 (0.18) | 0.96 (0.07) | 0.22 (0.04) | 66.0 | 35.6 |
| 87.7 μg/L | 93.3 | 60 | 4.58 (0.32) | 0.33 (0.05) | 0.60 (0.06) | 0.09 (0.08) | 58.0 | 8.5 |
- a SE = standard error.
| EmT50 (d) | |||
| Treatment/species | Males | Females | Adults collectively |
| Artificial | |||
| Control: C. riparius | 15.3 | 18.7 | 17.3 |
| 2.41 mg/L C. riparius | 15.2 | 18.7 | 16.8 |
| Natural | |||
| Control: C. riparius | 14.3 | 18.9 | 16.3 |
| 2.41 mg/L: C. riparius | 16.7 | 18.4 | 17.5 |
| Artificial | |||
| Control: C. tentans | 32.0 | 35.9 | 34.9 |
| 2.41 mg/L: C. tentans | 37.3 | − b | 39.7 |
| Natural | |||
| Control: C. tentans | 31.4 | 35.1 | 32.8 |
| 2.41 mg/L: C. tentans | 38.1 | 38.6 | 38.6 |
| Artificial | |||
| Control: C. riparius | 15.6 | 18.8 | 17.3 |
| 87.7 μg/L: C. riparius | 15.4 | 18.7 | 17.1 |
| Natural | |||
| Control: C. riparius | 16.0 | 19.0 | 17.8 |
| 31.6 μg/L: C. ripariusc | 15.1 | 18.5 | 17.4 |
| Artificial | |||
| Control: C. tentans | 37.0 | 39.5 | 38.7 |
| 87.7 μg/L: C. tentans | 36.1 | 37.0 | 37.2 |
| Natural | |||
| Control: C. tentans | 29.9 | 34.6 | 32.6 |
| 31.6 μg/L: C. tentansc | 37.1 | 39.8 | 38.6 |
- a Data shown represent the control and highest treatment concentration for each sediment.
- b Calculation of EmT50 was not possible.
- c Intermediate lindane concentration is shown because of high mortality at 87.7 μg/L, which did not allow calculation of EmT50.
Cadmium. In the cadmium-spiked sediments, no significant difference (p > 0.05) in the emergence times of C. riparius from control or treated artificial sediments was found. However, the EmT50s of males and of adults collectively (i.e., males and females combined) were significantly delayed (p < 0.05) at the highest exposure concentration in natural sediment compared with controls. For C. tentans, emergence of males, females, and adults collectively was significantly delayed at the highest treatment concentration (2.41 mg/L) in both sediment types.
The percentage emergence data (Fig. 1) revealed that the total number of C. riparius adults that emerged from each of the control and cadmium-spiked containers was in excess of 85%, which was significantly greater in all cases than the corresponding number for C. tentans. No significant difference (p = 0.627) in the percentage emergence of adult C. riparius was recorded in relation to cadmium exposure. In addition, the sex ratio of adults did not deviate markedly from the expected 1:1 relationship for this species. In contrast, the emergence of C. tentans failed to achieve the 70% acceptance criterion [23] for either controls or any of the treatment concentrations. The highest collective emergence of 43% was reached in the natural sediment control, and in general, emergence from the other treatments was only 20 to 30%. No significant difference (p > 0.05) in the numbers of C. tentans adults between treatments was noted for either sediment. A change in the sex ratio to a 2:1 relationship was noted in several treatments; however, no consistent pattern could be discerned.
Percentage emergence of Chironomus riparius and C. tentans adults after larval exposure to cadmium-spiked sediment. Data represent the mean of three replicates with 1 standard error. Art. = artificial sediment, Nat. = natural sediment, black bars = % of C. riparius emerged, white bars = % of C. riparius that are male, black diagonal-striped bars = % of C. riparius that are female, gray horizontal- and vertical-striped bars = % of C. tentans emerged, gray horizontal-striped bars = % of C. tentans that are male, gray vertical-striped bars = % of C. tentans that are female.
Lindane. In the lindane-spiked sediments, no significant difference (p ≥ 0.05) in the EmT50s of C. riparius adults between the different treatment groups was noted for males, females, or adults collectively. Emergence of C. tentans males, females, and adults collectively from the lindane-spiked artificial sediment occurred significantly earlier (p < 0.05) than in the control group. Similar analysis for the natural sediment confirmed a more predictable pattern, with the emergence of males, females, and adults collectively occurring significantly earlier in the control group than at any of the three lindane concentrations. Comparative analyses between sediments also confirmed that in general, the emergence of males, females, and adults collectively from lindane-spiked natural sediment was significantly delayed (p < 0.05) compared with that from artificial sediment.
As with the cadmium test, the percentage emergence of C. riparius exceeded that of C. tentans in all cases (Fig. 2). In artificial sediment, the emergence of C. riparius adults exceeded 90% in the control and the lindane-spiked replicates. As a result, no significant difference (p ≥ 0.05) was noted in the percentage of adults that emerged. However, in natural sediment, a significant reduction (P ≤ 0.001) in the number of adults was recorded at the two highest lindane concentrations. Emergence from the control and the lowest exposure concentration (16.6 μg/L) was greater than 90%, but this was reduced to 66% at 31.6 μg/L. At the highest concentration of 87.7 μg/L, only 5% of adults (three females) successfully emerged. Comparison between sediment types revealed that emergence was lower in natural than in artificial sediment spiked at 87.7 μg/L, but emergence was not significantly different (p ≥ 0.05) in the other corresponding treatments.
Percentage emergence of Chironomus riparius and C. tentans adults after larval exposure to lindane-spiked sediment. Data represent the mean of three replicates with 1 standard error. Art. = artificial sediment, Nat. = natural sediment, black bars = % of C. riparius emerged, white bars = % of C. riparius that are male, black diagonal-striped bars = % of C. riparius that are female, gray horizontal- and vertical-striped bars = % of C. tentans emerged, gray horizontal-striped bars = % of C. tentans that are male, gray vertical-striped bars = % of C. tentans that are female.
Adults of C. tentans failed to achieve the acceptable 70% emergence in either sediment control group, typically reaching only 50 to 60%. In artificial sediment, for example, the percentage of adults that successfully emerged was between 30 and 70%. Statistical analysis showed that emergence differed significantly (p < 0.05) between treatments, with multiple comparison analysis confirming that significantly fewer adults emerged at the lowest lindane concentration. Emergence of adults from natural sediment was between 50 to 60% in the control group and at lower exposure concentrations, but significantly fewer adults (15%) emerged from the replicates spiked at the highest concentration.
DISCUSSION
The results of this investigation show that both the growth and the emergence of C. riparius and C. tentans were affected by exposure to cadmium- and lindane-spiked sediments, but that the response criteria were influenced by the sediment type, choice of test species, toxicant, and duration of the experiment. The growth study illustrates several of these points, particularly the influence of sediment type and test species on the eventual result. Differential species sensitivity was noted in artificial sediment spiked with cadmium at 2.41 mg/L in which larvae of C. riparius had developed beyond the second instar, whereas larvae of C. tentans failed to do so. This difference is reflected by the C. riparius larvae exposed at this concentration reaching 61% of control wet weight, whereas the corresponding value for C. tentans was only 10%. Further evidence for a difference in sensitivity is provided in natural sediment, in which 100% mortality among C. tentans occurred at the higher exposure concentrations (4.8 and 9.6 mg/L), but in which C. riparius survival was relatively unaffected. Differences in toxicity were also noted between sediments, with cadmium-spiked natural sediment exerting a greater toxic effect than the corresponding treatment group in artificial sediment. This is shown by the significantly greater growth of C. riparius at 2.41 mg/L in the artificial sediment, with these larvae achieving 61% of control wet weight, compared with natural sediment, in which larvae only achieved 9% of the control weight. Additional evidence for differential toxicity between sediments is provided by the reduced survival of C. tentans in natural sediment.
Sediment geochemistry affects toxicant bioavailability [24], and our results may reflect differences in the physical and chemical characteristics of the two sediments, with a greater amount of metal being available to larvae in the natural sediment. However, assuming that the bioavailable fraction of metal for benthic species resides in the pore water [25, 26], the greater level of growth inhibition and mortality associated with natural sediment would not be predicted from the results of chemical analysis. Cadmium concentrations in pore water were consistently greater in the artificial sediment, and the aqueous LC50 of 0.7 mg/L for C. riparius and C. tentans, as determined previously in our laboratory (M.M. Watts and D. Pascoe, unpublished data), was not exceeded in either aqueous fraction. That the greatest difference in cadmium concentration, which is consistent with the observed results, arose in the respective sediment loads suggests that the sediment itself was the primary route of exposure, presumably via ingestion or tissue absorption. Similar findings have been reported by Suedel et al. [27], in which the toxicity of copper to C. tentans larvae corresponded to sediment concentration as opposed to that associated with overlying or pore water. Bioavailability cannot be predicted purely on the basis of whole sediment concentrations [24], but this does not discount it as a source of toxicity. The results of this investigation seem to support this view.
Several points should be considered, however, in addition to those already outlined that could potentially influence the interpretation of results from the two sediments, most notably the relationship between sediment particle size and the contribution of larval gut contents to final weight. Brooke et al. [28], for example, showed that 10% of the final dry weight of C. tentans larvae was accounted for by inorganic gut contents, whereas Sibley et al. [29] found that 7 to 59% of the final dry weight for C. tentans exposed in different sediments was composed of gut contents. Sibley et al. concluded that those sediments with a mean particle size distribution of less than 306 μm, which presumably can be easily ingested by larvae, posed the greatest risk of erroneous data interpretation. To correct for these variables, Sibley et al. proposed the determination of ash-free dry weight, and they found that the number of statistically significant outcomes was reduced after such correction. Because the natural sediment used in the present study was composed of a larger number of finer particles than the artificial sediment, the possibility exists that more would have been ingested by the larvae—possibly accounting for the generally higher weight of control larvae and the perceived increase in toxicity associated with the natural sediment. These modifying factors need to be considered when comparing larval growth between artificial and natural sediments, but they are not a consideration when comparing species response in the same sediment.
In contrast to the cadmium study, the growth of larvae exposed to lindane-spiked sediments revealed no difference in either the wet or the dry weight of larvae between sediment types, except for C. riparius exposed at the intermediate concentration of 31.6 μg/L. However, the percentage loss of weight compared with the control group showed that differential effects on growth were occurring between sediment types. At the highest exposure concentration of 87.7 μg/L, C. riparius larvae in artificial sediment reached 74.5% of the control weight, whereas in the natural sediment, the figure was 58%. Similarly, for C. tentans, 27.4% of the control weight was noted for animals in the artificial sediment, with larvae in the natural sediment only achieving 8.5%. This suggests that percentage weight loss may provide a useful additional tool when assessing the differences in toxicity between different sediments. On the basis of the cadmium study, increased lindane toxicity in natural sediment may have been predicted. However, the increased bioavailability of a particular chemical in one sediment does not necessarily equate to a similar effect for all chemicals, because the nature of the contaminant itself also plays an important role [30].
Larval growth in lindane-spiked sediments confirms the greater sensitivity of C. tentans, because the mean wet weight was reduced significantly in both sediments at all lindane concentrations. For C. riparius, this did not occur at the lower exposure concentrations in the artificial sediment. In addition, the percentage weight loss in relation to controls was greater for C. tentans than for C. riparius at each exposure concentration.
Each spiking concentration of cadmium and lindane adversely affected the larval growth of both test species, but a similar affect was not noted in the emergence study. Toxicant-related effects on this endpoint were primarily noted at the higher concentrations in natural sediment, with little effect on adult numbers or emergence times in either species being recorded at the lower concentrations or in artificial sediment. Spiking the test sediments with cadmium did not affect the number of adults emerging for either C. riparius or C. tentans in relation to their respective controls; however, emergence times were affected by cadmium spiking.
In the lindane study, delayed emergence and reduced numbers of adults for both C. riparius and C. tentans were noted. Reduced numbers of C. riparius adults emerging from natural sediment, presumably because of larval mortality, was found at lindane concentrations of 31.6 and 87.7 μg/L. However, this does not reflect the results of the growth experiment, in which larval mortality was not associated with these concentrations. This apparent increase in sensitivity may be attributable to one or more factors, such as the greater susceptibility of first-instar larvae in relation to second-instar larvae [31] used for the growth test. A second contributory factor could be the increased exposure time, which Suedel et al. [27] identified as producing an increased sensitivity of C. tentans exposed to copper-spiked sediment.
The control survival of both species in the growth test was 100%. Emergence test data for C. tentans are problematic, however, because percentage emergence in controls failed to achieve the 70% acceptance criterion [23] in either the cadmium or the lindane experiments. The “typical” control emergence of this species from long-term tests has been reported to range from 60 to 70%, despite the fact that after a 20-d exposure, larval survival was in excess of 90% [12]. This suggests that death was occurring independently of toxicant stress and during the latter stages of the test, corresponding to larval development through the fourth instar leading to eventual pupation and eclosion. As a consequence, the level of confidence in the data associated with C. tentans emergence is low and, thus, of limited use for an assessment of toxicity.
Variable emergence among C. tentans, often to levels lower than the 70% criterion, can be attributed to several factors in addition to toxicity, such as food levels and sediment particle size [3, 12, 29]. With this in mind, the poor performance of C. tentans recorded in this investigation may be attributable to the test conditions. Several authors [3, 32] have noted that a minimum dry larval weight of 0.5 to 0.7 mg/larva must be reached by C. tentans for emergence to occur. Weights were not determined in the emergence test, but the results from the growth test, in which larvae reached this threshold weight (most were still only third instar), would seem to discount the sediments themselves as being unsuitable and, therefore, responsible for the poor performance of C. tentans. However, additional factors such as competition for space and food between larvae or cannibalism of smaller larvae cannot be discounted in explaining the low survival and subsequent emergence level.
Having accepted that factors such as those described may have influenced the emergence of C. tentans, the emergence of this species remains generally poor, with variable numbers of adults being noted in replicates of the same treatment. Perhaps as a concession to this variability, recent discussions by the U.S. Environmental Protection Agency have led to a proposal for lowering the criterion of C. tentans emergence to 50%.
CONCLUSION
This investigation demonstrated that sediment toxicity is a complex issue, with differences in sediment type, toxicant, and test duration affecting both organism response and subsequent data interpretation. In addition, the assumption that C. riparius and C. tentans are similarly effective as indicators of sediment toxicity is now brought into question. For sediment assessment in tests of less than 10-d duration or those not incorporating the latter stages of the life cycle, C. tentans is perhaps more suitable, because its sensitivity to both cadmium and lindane was greater than that of C. riparius in the growth study. However, for tests of longer than 10-d duration or those relying on the emergence of adults, the situation is complicated by the poor performance of C. tentans. That the emergence of this species was below the acceptance criteria, which are currently in the process of being reduced further, must give some cause for concern regarding the reliability of results. This problem can be partially solved through increased replication, but that C. riparius presented no such difficulties suggests this species is perhaps more appropriate for tests of this nature. In an ideal situation, sediment assessment should be performed with both species, as in the present study, thereby providing a balance of sensitivity (C. tentans) with reliability (C. riparius). These findings should help to improve the existing test protocols for the assessment of sediment toxicity and, ultimately, to improve the level of understanding regarding this problem.
Acknowledgements
The authors thank Zeneca Agrochemicals for financial support and the supply of natural sediment, M. O'Reilly and O. James for technical assistance, and the two reviewers for their constructive comments on the manuscript.
