Z’ozicologyLetters, 64/65 (1992) 511-517 0 1992 Elsevier Science Publishers B.V., All rights reserved 03784274/92/$5.00

511

Hazard assessment in freshwater ecosystems N.O. Crossland Biology Department,

University ofMississippi,

0xforc-J MS (USA)

Kq words: Expected environmental concentrations; No observed effect concentrations; Hazard assessment; Freshwater ecosystems

SUMMARY Hazard assessment of chemicals in freshwater environments depends on comparing concentrations that are expected to occur in water and sediment, i.e. expected environmental concentrations (EEC), with those that are estimated to have no biological effects, i.e. the no-observed effect concentrations (NOEC). The difference between these two estimates is the margin of safety. The EEC can be estimated from data for chemical release rates, physicochemical properties and environmental parameters that affect transport and transformation. The NOEC can be estimated from the results of toxicity tests using aquatic plants, invertebrates and fish. When making these estimates it may be necessary to extrapolate from relatively limited laboratory data to the real world. Inevitably, this involves some degree of uncertainty. Such uncertainty can often be resolved by carrying out controlled field tests, using small, outdoor enclosures (microcosms), relatively large, outdoor ponds (mesocosms) and experimental streams. In this paper the advantages and disadvantages of various experimental approaches and systems will be reviewed.

INTRODUCTION

Historically, many freshwater environments have been used as convenient depositories for municipal and industrial wastes. Such practices have led to severe degradation of the quality of many rivers and lakes in developed countries of the world. The problems caused by such pollution have been exacerbated by increasing demands on limited supplies of fresh water. In many countries river water is re-cycled several times before being discharged to the sea. Thus, it is not surprising that most regulatory authorities are particularly concerned to prevent the pollution of freshwater environments by toxic chemicals. Beginning in the 1970s hazard assess-

Correspondence to: N.O. Crossland, The Shingles, Spa Esplanade, Herne Bay, Kent CT6 8EP, UK

512

ment procedures have evolved and are now applied by regulatory authorities to determine the nature and magnitude of hazards that may result from the release of chemicals into the environment. These procedures depend on comparing chemical concentrations that are expected to occur in the environment, the Expected Environmental Concentrations, or EECs, with those that are estimated to have no biological effects, the No Observed Effect Concentrations, or NOECs. Differences between EECs and the NOECs are margins of safety. SEQUENTIAL TESTING

Hazard assessment procedures incorporate a series of steps or tiers progressing from relatively simple to more complicated laboratory tests and finally to field tests and environmental monitoring. An initial or screening stage is based on physicochemical properties, rates of degradation and acute toxicity tests using a single species of fish, a crustacean (Daphnia sp.) and a species of aquatic plant (a unicellular alga). An intermediate stage involves toxicity tests using a wider range of organisms, tests to assess sub-lethal effects, bioaccumulation studies and assessment of persistence in water and sediment. An advanced stage of hazard assessment may involve testing in outdoor ponds, plastic enclosures in lakes or in artificial streams to assess the fate and effects of chemicals under conditions that closely simulate the natural environment. At an early stage in the hazard assessment procedure, uncertainties associated with estimates of the EECs and the NOECs may be relatively great and the associated confidence intervals may be relatively wide. As testing continues and further data are generated the uncertainties become less and confidence intervals become narrower. Testing may continue until the margin of safety is acceptably wide, or can be defined with suffkient confidence. THE EXPECTED ENVIRONMENTAL CONCENTRATION (EEC)

The EEC can be estimated from data for chemical release rates, together with patterns of use, physicochemical properties and environmental parameters that may affect the rates of dispersion and degradation. Mathematical models can be used to couple physicochemical properties with environmental parameters and thus to estimate the EEC for specified environments. The reliability of such estimates depends, to a great extent, on the reliability of the inputs to the models, particularly on the data for chemical release rates and physicochemical properties. Data for vapor pressure, water solubility and the octanol/water partition

513

coefficient can be used to make an initial assessment of the likely distribution of a chemical between water, air, sediment and biota. The ratio of vapor pressure to water solubility is Henry’s constant, H, which can be used to estimate the tendency for a chemical to volatilize from water into air. If H is much less than 1.0 a chemical will tend to remain in water, whereas if H is much greater than 1.0 it will tend to escape into the atmosphere. Values for H vary from around 1 x 10e5 for relatively involatile, water soluble chemicals such as the herbicide 2, 4-D up to IO4 for volatile solvents. Even involatile organic chemicals can have high values forH, if they also have low values for water solubility. For example, the vapor pressure for DDT is 2.63 x lOa Pa and its water solubility is 9.36 x lOa mol m3, from which H may be calculated as 2.84 Pa m3 mol-r. Therefore, DDT, despite its very low vapor pressure, tends to escape from water into the atmosphere. The octanol/water partition coefficient, I&,,, can be determined using a shake-flask technique, in which the chemical being tested is added to a flask containing octanol and water. After shaking this mixture the two phases are allowed to separate and chemical concentrations in each phase are determined by appropriate methods of chemical analysis. The ratio of these two concentrations gives K,,. Alternatively, Kow can be determined using high performance liquid chromatography (HPLC). Values obtained for K,, can be used to estimate the distribution of a chemical between water and sediment at equilibrium. Chemicals with a high value for K,,, are more likely to become associated with sediments than those with a low value. I&,,. also provides a good estimate of the relative solubility of organic compounds in water and animal fats and is therefore widely used to estimate the potential for bioaccumulation in fish and other aquatic animals. THE NO OBSERVED-EFFECT-CONCENTRATION

(NOEC)

On the basis of toxicity tests, the NOEC can generally be estimated with greater confidence than the EEC. Even so, the NOEC is based on biological data that are themselves subject to statistical variability. Moreover, there are uncertainties in extrapolating from acute to chronic effects, from one species to another and from the results of laboratory toxicity tests to biological effects in real-world environments. To allow for these potential sources of error, “uncertainty factors” are often applied to the data to estimate the NOEC. For example, an OECD Working Group 111suggested that an uncertainty factor of 1000 should be applied to data for acute toxicity to estimate the NOEC, a factor of 100 to relatively limited data for chronic toxicity and a factor of 10 to more extensive and reliable date for chronic toxicity. If data from field tests are available an uncertainty factor of only 1 might be appropriate.

514

Juvenile fish of various species, e.g. rainbow trout, carp, bluegill sunfish, fathead minnows, may be used for acute toxicity tests during early screening. They are exposed to a logarithmic series of chemical concentrations for a period of 96 h and mortality is determined at intervals of 24 h. The results are expressed as the 96 h LC50, i.e. the concentration of chemical that is lethal to 50% of the fish population after exposure for 96 h. A similar procedure is used to evaluate acute toxicity to Daphnia sp. This planktonic crustacean is relatively easy to rear in the laboratory and it reproduces parthenogenetically. This feature of its life-cycle ensures that genetic variation between individuals in toxicity tests is minimal. Newly-born individuals, less than 24 h old, are exposed to chemicals for a period of 48 h and results are expressed as the 48 h effective concentration, EC5o. Daphnia is relatively sensitive to many environmental contaminants, its use in toxicity tests is well established and therefore it has become a key organism in hazard assessment procedures. Toxicity tests with unicellular algae are carried out using cultures maintained in flasks of nutrient solutions. The numbers of algal cells in each flask are counted at the start of a test and again after 72 h exposure to a series of chemical concentrations. The growth rate and biomass of the populations of algal cells are calculated and the results expressed as the 72 h EC 50, i.e. the concentration that inhibits growth rate or biomass production by 50% after 72 h exposure. The results of these three acute toxicity tests may be used to make a preliminary assessment of possible hazard to freshwater ecosystems. The NOEC may be estimated using appropriate uncertainty factors, as described previously. This value can then be compared with an estimate of the EEC to give an estimate of the margin of safety. If this is sufficiently wide no further tests may be required. If not, further testing may be indicated. Chronic or long-term toxicity tests may be carried out using various species of fish and Daphnia. In these tests sensitive toxicity end points, such as growth and reproduction, are evaluated. From the results of such tests it is possible to define the NOEC with greater accuracy than on the basis of acute toxicity tests. Whole life-cycle tests with fish are sometimes carried out using species that complete their whole life-cycle, from egg to egg, during a period of about 9 months. However, it has been shown that the embryonic and early larval stages of the fish life-cycle are usually the most susceptible and therefore a shortened test has been developed which lasts for only 30 days. This embryo-larval test is most often carried out using the fathead minnow, Pimephales promelas, a species native to North America. Batches of eggs are obtained from pairs of spawning adults and placed in chambers containing water and chemical. Water and chemicals are continually provided to the chambers under “flow-through” conditions. Hatching of the eggs, abnormal development and growth rate of the larvae are monitored.

515

Alternatively, effects on fish growth rates can be evaluated using juvenile rainbow trout. Provided food and space are not limiting, the growth rate of this species is constant, relatively fast and predictable from the fry stage to the onset of maturity. Fish are individually marked and exposed in batches to a series of chemical concentrations. They are weighed and measured at the start of the test and again after exposure for 14 and 28 days [21. Growth rate is usually the most sensitive end point in chronic toxicity tests with fish while reproduction is usually the most sensitive end point in chronic toxicity tests with Duphnia. With this species tests generally last for 21 days, during which time three generations of offspring will be produced. Newly born Duphnia are placed individually in beakers containing water and chemical. Every 24 h the Duphniu are transferred to freshly-prepared beakers. Effects on mortality and reproduction are monitored daily. The organisms reach maturity after 7-10 days and a single Daphniu may produce about 100 offspring during the 21-day test. Bioaccumulation tests may be required for chemicals where Kowis greater than 1000. The bioaccumulation factor, or BCF, is determined as the ratio of chemical concentration in fish: chemical concentration in water, after the fish have been exposed long enough for chemical concentrations in fish and water to have reached equilibrium. For chemicals with a high potential for bioaccumulation, e.g. some organochlorines and heavy metals, an exposure of several months or even years may be required to reach equilibrium. To overcome the problems associated with such long exposures, a shortened method for determining BCF has been developed. This depends on evaluating the rates of uptake and depuration of chemicals in fish. The ratio of these rates is equivalent to the BCF at equilibrium. Batches of fish are exposed to chemicals, generally under flow-through conditions. Concentrations of chemicals in water and fish are determined at various times during the “uptake” phase of the test. Fish are then transferred to tanks containing water but no chemical and concentrations of chemical in the fish are determined at various times during the “depuration” phase. On the basis of data obtained from acute and chronic toxicity tests and bioaccumulation studies it may be possible to estimate the NOEC with sufficient confidence. On the other hand, if uncertainties still exist concerning the acceptable margin of safety, studies may be carried out under controlled field conditions. FIELD TESTS IN FRESHWATER

ENVIRONMENTS

Uncertainties arising from preliminary hazard assessments may be many and varied. They may concern the dispersion and degradation of a chemical, bioaccumulation, its spectrum of toxicity, its potential for indirect effects, or

516

any combination of these factors. Field tests of various kinds may be carried out to investigate questions that cannot be resolved using standard laboratory tests. Such tests cannot be carried out in natural aquatic environments because of the lack of control of environmental variables, the need to replicate experimental systems and the risks of causing environmental damage. ‘Field tests are therefore carried out using aquatic mesocosms, which have been defined as bounded and partially enclosed outdoor experimental units that closely simulate the natural environment [31. For regulatory purposes the U.S. Environmental protection Agency (EPA) has issued guidelines defining the objectives and scale of mesocosm studies [41. Those carried out for pesticide registration may have two regulatory objectives. The first is to negate presumptions of unacceptable risks based on comparisons between estimates of the EEC and the NOEC. The second is to provide descriptive information on the duration and magnitude of adverse impacts. The recommended size of mesocosms is 0.1 ha and 12 of them are required for a test that involves three replicates of three treatment levels plus a control. The results of such tests have provided useful information for hazard evaluation [5,6]. However, they are very costly and labour-intensive and it has been suggested that equally valid information can be obtained using smaller systems. Evidence to support this view is somewhat limited but vigorous research efforts are being directed towards investigating this key question of scale in mesocosm tests. For example, the fate and effects of the pyrethroid insecticide cyfluthrin were investigated in three 75 m2 natural ponds and compared with fate and effects in three, 3 m2 artificial ponds. Very similar results were reported from both systems while use of the smaller systems resulted in considerable savings in costs. At a workshop convened by SETAC (Society of Environmental Toxicology & Chemistry) the use of miniature mesocosms was reviewed by a group of 39 scientists [81. These systems, approx. 5-10 m3 in size, are miniature versions of the 0.1 ha mesocosm ponds recommended in the EPA guideline 141.The workshop concluded that results from such miniature mesocosms would normally be sufficient to resolve the issues that normally trigger a larger and more expensive mesocosm study. However, further validation of these systems will be required before they can be adopted for regulatory purposes. REFERENCES 1

2

Organization for Economic Co-operation and Development (1989). OECD Environment Monographs No. 26. Report of the OECD Workshop on Ecological Effects Assessment, OECD, 2, rue Andre-Pascal, 75775 Paris Cedex 16, France. Crossland, N.O. (1988) In: W.J. Adams, G.A. Chapman and W.G. Landis (Eds.), Aquatic Toxicology and Hazard Assessment: 10th Vol., American Society for Testing and Materials, Philadelphia, pp. 463-467.

517 3 4 5

6

‘7 8

Odum, E.P. (1984) The mesocosm. Bioscience 34,558562. Touart, L.W. (1988) Hazard Evaluation Division Technical Guidance Document. US Environmental Protection Agency, Washington D.C. Report No. EPA/540/0988-035. Webber, E.C., Deutsch, W.G., Bayne, D.R. and Seesock, W.C. (1992) Ecosystem-level testing of a synthetic pyrethroid insecticide in aquatic mesocosms. Environ. Toxicol. Chem. 11,87-105. Fairchild, J.F., La Point, T.W., Zajicek, J.L., Nelson, M.K. et al. (1992) Population-, community- and ecosystem-level responses of aquatic mesocosms to pulsed doses of a pyrethroid insecticide. Environ. Toxicol. Chem. 11,1X-129. Heimbach, F., Pflueger, W. and Ratte, H.-T. (1992) Use of small artificial ponds for assessment of hazards to aquatic ecosystems. Environ. Toxicol. Chem. 11,27-34. SETAC-Resolve Workshop on Aquatic Microcosms for Ecological Assessment of Pesticides (1992) SETAC Foundation for Environmental Education, Inc. 1010 North 12th Avenue, Pensacola, FL 32501, USA.

Hazard assessment in freshwater ecosystems.

Hazard assessment of chemicals in freshwater environments depends on comparing concentrations that are expected to occur in water and sediment, i.e. e...
514KB Sizes 0 Downloads 0 Views