CADMIUM AND LEAD SOURCES FOR BENTHIC INVERTEBRATES INFERRED FROM IN SITU EXPERIMENTS

André Tessier and Landis Hare, INRS-Eau, Université du Québec, C.P. 7500, Sainte-Foy (Qc), G1V 4C7, Canada.

Corresponding author: atessier@inrs-eau.uquebec.ca

ABSTRACT. Current strategies for metal-contaminated sediment management assume that benthic animals take up their metals from sediments. The source of metals to benthic animals is, however, not obvious, since they inhabit sediments but irrigate their burrows with oxygenated overlying water. We conducted in situ experiments in two lakes to determine the relative importance of Cd and Pb in the sedimentary compartment versus Cd and Pb in the overlying water compartment for accumulation of these metals by benthic animals. There are indications that, for most of the taxa studied, individuals living in the two lakes took up their Cd (and perhaps their Pb) mainly from the overlying water compartment.

INTRODUCTION. Many bottom-dwelling animals burrow into the anoxic zone of sediments at depths were the bulk sediment is anoxic and contains sulfide. It is generally assumed that these benthic animals are exposed to metals in sediments and porewaters and, accordingly, sediment quality criteria have been set forth to protect them. However, most benthic animals, even if they live in anoxic sediments cannot tolerate anoxic conditions for extended periods and thus irrigate their burrows with oxygenated water from above the sediment surface (Aller, 1982). In doing so, animals create microenvironments that are different from the bulk sediment in which is located their burrow. Conditions in these microenvironments may resemble more those existing above the sediment-water interface than those prevailing in the bulk sediment. Given their behavior, benthic animals could take up metals from water and particles drawn through their burrows as well as from the anoxic sediment and interstitial water in which their burrows are situated. Thus a fundamental question to ask is: are metal concentrations in benthic animals related more to metal concentrations in the overlying water or in the sediment? These two compartments are considered herein in a broad sense: the sediment compartment refers to any or all of pore waters, sediment particles and specific sediment components (e.g., AVS, organic matter, etc.); the overlying water compartment refers to the overlying water itself and to those food particles whose metal concentrations are directly related to dissolved metal concentrations in the overlying water. We conducted field experiments to determine the relative importance of the sediment and overlying water compartments as pools from which freshwater benthic animals accumulate Cd and Pb.

METHODS. We assumed that the total metal (M = Cd or Pb) concentration, [M_animal]_total, accumulated by a benthic animal is the sum of metal it accumulates from the overlying water compartment, [M_animal]_o.w., and of metal that it accumulates from the sediment compartment, [M_animal]_sediment, that is:

[M_animal]_total = [M_animal]_o.w. + [M_animal]_sediment (1)

The experimental approach we used to quantify the relative size of these two contributions to total animal metal concentrations involved creating a gradient of sediment metal concentrations in two lakes (L. St. Joseph and L. Laflamme), while leaving metal concentrations in the overlying waters unaltered. Lake St. Joseph is a 5-km², heavily cottaged water body whereas L. Laflamme is a small (0.6 km²) headwater lake; the two lakes are located on the Canadian Shield, are heavily forested but they support different benthic communities. Independent experiments were carried out for Cd and Pb at a littoral station in each of the two lakes.

In practice, the approach entailed collecting large quantities of sediment from the lake bottom to which we added various quantities of either Cd or Pb to attain nominal metal to AVS (Acid Volatile Sulfide) ratios both below and above 1. The spiked sediments were then returned by divers to the lake bottom in the early spring in plastic containers (surface area 0.09m²) that were left open to lake water for about 1 year to allow colonization by a natural community of benthic organisms. In the next spring, before emergence of most insect species, porewater and overlying water were sampled with peepers from a container of each treatment level and profiles of pH and Ca, Cd, Fe, Mg, Mn, Pb, SS(-II), SO₄ and inorganic and organic carbon concentrations were determined, using analytical methods described elsewhere (Warren et al., 1998). Sediment cores were collected from the same containers, sliced into 1-cm sections that were analysed for AVS, SEM (Simultaneously Extracted Metals), total Cd, Fe, Mn, Pb and organic carbon. The remaining containers (n=7 per treatment level) were covered and brought to the lake surface. Benthic animals were isolated by sieving, sorted according to type, counted, placed in lake water to evacuate their gut contents, dried, digested and analysed for their Cd or Pb content.

Additional experiments were also carried out in the autumn and in the spring to estimate invertebrate residence times in the containers and metal uptake rates. Containers filled with lake sediments contaminated with either Cd or Pb were placed on the lake bottom where they were left open to allow colonization. The containers were recovered at various time intervals from 4 to 56 d for animal enumeration and measurement of their Cd and Pb content. Residence times were estimated as the inverse of the proportion of animals leaving the containers per unit time, as described in Hare (1995).

RESULTS AND DISCUSSION. At sediment depths below 2 cm, sedimentary Cd and Pb concentrations measured in the containers at the end of experiments were close to the expected values. In contrast, Cd and Pb values in the upper sediment layers were lower due to upward fluxes of dissolved metals and to the exchange of Cd- and Pb-poor particles from outside of the experimental containers with Cd- and Pb-rich particles inside of the containers during resuspension events. The concentrations of dissolved Cd generally increased with Cd treatment level, but this trend was not consistently the case for Pb. The AVS model (Di Toro et al., 1992) assumes that metals (such as Cd and Pb) that form less soluble solid sulfide compounds will replace the more soluble metal (such as Fe and Zn) in solid sulfides present in sediments. According to this model, porewater Cd²⁺ and Pb²⁺ concentrations should be low in our containers as long as the Cd and Pb concentrations added to the containers are lower than the AVS concentration and then should increase sharply when the AVS concentration is exceeded. Cadmium appears to conform to this behavior (Fig. 1a) whereas Pb does not (Fig. 1b).

The invertebrate community in our experimental containers was dominated by insects in both lakes. The only common non-insects were tubicifid and lumbricid oligochaete worms. Most of the taxa for which we had sufficient numbers to estimate residence times remained in the experimental containers for from 2 to 11 weeks, and temporal variations in animal metal concentrations suggest that steady state was reached. In both lakes, no taxon was eliminated by

Figure 1. Scatter plot of pore water free Cd (a) and free Pb (b) activities versus [SEM-Cd]-[AVS] or versus [SEM-Pb]-[AVS]. The vertical dashed line represents the point at which the AVS model predicts that increases in [Cd²⁺] and Pb²⁺] should occur.

the added Cd and Pb and there was no difference in the mean densities of the communities collected from the control containers (no added metals) and those with the highest Cd (7.7 µmol g^-1 in L. St-Joseph and 2.8 µmol g^-1 in L. Laflamme) and Pb (4.5 µmol g^-1 in L. St-Joseph and 10.3 µmol g^-1 in L. Laflamme) concentrations, which suggests that the presence of Cd and Pb in sediment had little influence on animal densities. Likewise, there was no significant difference in the density of most individual taxa between the extremes of our Cd or Pb treatment levels. Exceptionally, the densities of tubicifid oligochaetes were lower at high Cd and Pb concentrations in L. St. Joseph, a difference that could be explained by their behavior. The lack of Cd and Pb effects at the community level is surprising given the fact that the concentrations of sedimentary Cd and Pb used in the experiments largely exceed sediment quality objectives of many countries (0.005 – 0.01 µmol g^-1 for Cd and 0.12 – 0.48 µmol g^-1 for Pb; de Vries and Bakker, 1996) and the Cd and Pb concentrations reported to occur in recent sediments of many metal polluted lakes. These results suggest that most benthic animals are not exposed to the metals in the bulk sediment.

In our experiment (various metal levels in otherwise similar sediment among treatments), the concentration of metal that an animal takes up from the sediment compartment should depend on the metal concentration of the sediment, [M_sediment] , that is :

[M_animal]_sediment = F_sediment [M_sediment] (2)

where F_sediment is a proportionality constant specific to an animal in a given lake. Combining Eqns 1 and 2 gives:

[M_animal]_total = [M_animal]_o.w. + F_sediment [M_sediment] (3)

Figure 2. Scatter plot of [Cd_animal]_total versus sediment total Cd for Sialis velata from L. St. Joseph (a) and Procladius spp. from L. Laflamme (b). Dotted lines represent the 95% confidence interval about the regression lines.

which is the equation of a straight line. Cadmium concentrations in most taxa were directly related to those in sediment in experimental containers; Figure 2 shows two examples. This trend is in disagreement with the AVS model that predicts that Cd bioaccumulation should occur only when [SEM-Cd] > [AVS]. Animals accumulated from 10X to 100X less Pb than Cd for identical metal concentrations added to the sediment. Given the low Pb concentrations measured in the animals, the relationships between Pb in animals and Pb in sediment is not as clear as for Cd. Using the estimates of F_sediment for Cd for each taxon, we determined in the following manner the relative importance of the overlying water and sediment compartments as cadmium sources for animal living outside of our experimental containers. In Eqn 2, we substituted our experimentally-derived values of F_sediment and the value for [Cd_sediment] measured in the environs to calculate [Cd_animal]_sediment for each animal. These latter values were substituted in Eqn 1, along with those for the total Cd concentrations in animals from the environs, [Cd_animal]_total, to calculate Cd taken up from the overlying water compartment by animals in the environs, that is, [Cd_animal]_o.w.. By comparing the magnitudes of [Cd_animal]_o.w.with those of [Cd_animal]_total we estimated the percentage of the Cd in animals that came from the overlying water compartment. All insect taxa took up the majority of their Cd from the overlying water compartment, in spite of the fact that they live in the sediment. The water column compartment was of similar importance as a Cd source for taxa common to our two study lakes. In contrast, tubificid oligochaetes took up100% of their Cd from the sediment compartment. This difference between these two invertebrate groups is likely a consequence of differences in their behavior: most insects irrigate their burrow whereas tubificid oligochaetes do not. These results on the importance of the overlying water compartment for metal accumulation to benthic animals contradict the assumption of the AVS model that benthic organisms accumulate their metal from exposure to the interstitial water. Our results also suggest that sediment quality criteria might not be adequate

to protect all benthic animals.

REFERENCES

Aller R.C. (1982) In: Animal-Sediment Relations, P.L. McCall and M.J.S. Tevesz [Eds], Plenum Press, pp. 53-102.

de Vries W, Bakker DJ (1996), Manual for calculating critical loads of heavy metals for soils and surface waters. Preliminary guidelines for environmental quality criteria, calculation methods and input data. DLO Winand Staring Centre, report No 114.

Di Toro DM, Mahony JD, Hansen DJ, Scott KJ, Carlson AR, Ankley GT(1992), Environ. Sci. Technol. 26 : 96-101.

Hare L (1995), J. North Amer. Benthol. Soc. 14 : 315-323.

Warren L. Tessier A, Hare L (1998), Limnol. Oceanogr. 43 : 1442-1454.