Paracelsus' Dictum, addressing human toxicology, forms the basis for both modern medicine and ecotoxicology. The Dictum can be paraphrased as ‘The right dose makes the poison.’ What this means in terms of ecotoxicology, is that for toxicity to occur, an organism must be exposed to a contaminant and at the correct concentration. Effectively, this Dictum also separates pollutants (toxicity occurs) from contaminants (toxicity does not occur). However, as presently constituted, the Dictum neither includes nor considers two critical modifying factors, for example in the sedimentary environment: bioavailability and realistic exposure scenarios. Too many scientists, regulators and stakeholders are, explicitly or implicitly, applying the Dictum directly, without modification. Such literal interpretations have resulted in incorrect assessment of, for instance, the risk of polyaromatic hydrocarbon phototoxicity; the relative importance of pore water, overlying water and dietary exposure routes; biological tolerance mechanisms (avoidance, acclimation, non-genetic adaptation, genetic adaptation, metabolism); and secondary (indirect) toxicity. To ensure that the emphasis is on pollutants, not contaminants, the Dictum needs to be rephrased as follows and adhered to: ‘All substances are poisons; there is none which is not a poison. The right dose of a bioavailable substance, administered under realistic exposure conditions, differentiates a poison.’
Paracelsus (1493–1541) provided the basis for both human and ecological toxicology with his Dictum: ‘All substances are poisons; there is none which is not a poison. The right dose differentiates a poison. . .’. Basically this Dictum states that, for toxicity to occur, three factors are necessary: an organism, contaminant exposure, and the correct dosage (Figure 1). The Dictum correctly emphasizes the internal dose, not the external concentration. Although historically ecotoxicology has focused on external concentrations in aquatic systems, the emphasis is shifting to internal doses (Jarvinen and Ankley, 1999). The Dictum also effectively separates contamination from pollution, another important distinction. Contamination does not necessarily result in toxicity, and only comprises a substance out of place or present in excess. In contrast, pollution results in toxicity as a result of contamination causing adverse biological effects. All pollutants are contaminants, but not all contaminants are pollutants (Chapman, 2001).
However, the Dictum does not consider, for example in sediments, either bioavailability or realistic exposure scenarios. Effectively Paracelsus' Dictum only deals with hazard (the possibility of harm), not risk (the probability of harm). The purpose of this paper is to revise the Dictum to address risk, using examples derived primarily from studies with contaminated sediments. This revision is not simply semantic but rather embodies a necessary change in the mind-set of environmental scientists, regulators and other stakeholders.
Although it is the bioavailable fraction of contaminants in the environment that is of concern, and which can result in toxicity, typically this fraction is not measured. For instance, in sediments total concentrations of metals are measured. While measurements of the dissolved fraction of metals in the water column provide better indications of bioavailability than measurements of total concentrations, the dissolved fraction does not equate directly to the bioavailable fraction (Meyer, 2002). With respect to metals and organics, bioavailability is typically reduced with partitioning from the water column to the sediments; in general, contaminants are less likely to be toxic when bound to the sediments than when they occur in the water column (Chapman, 1999). A return to the water column from the sediments via physico-chemical processes or disturbance of the sediments (see below) is certainly a possibility, but typically bioavailability will be less than when the contaminants first entered the aquatic environment (Chapman et al., 2003). Sediment contaminants, though present at elevated concentrations, may not be bioavailable due to either their chemical state, or the nature of the sediment matrix. Paint chips, lead shot, tar balls, metal ore, soot, pitch or coke globules will result in increased sediment contaminant concentrations (e.g., metals, polynuclear aromatic hydrocarbons (PAH)), but limited bioavailability of those contaminants (Wenning and Ingersoll, 2002). Similarly, sediment matrices consisting of black carbon, peat or wood chips will limit the bioavailability of contaminants in those sediments (Jonker and Koelmans, 2002).
Further, biological mechanisms may limit the toxic effects of contaminants. Such mechanisms include acclimation (Weis, 2002), non-genetic adaptation (Millward and Klerks, 2002), genetic adaptation (Barata et al., 2002), and contaminant metabolism (Lovell et al., 2002). Genetic adaptation is not uncommon (Hulburt, 2002), and though genetic adaptation to particular contaminants may result in susceptibility to other stressors such as disease (Kammenga et al., 2001), the bioavailability of the original contaminants is reduced. In addition, infaunal organisms such as marine polychaetes can enhance the degradation (i.e., limit the bioavailability) of at least some organic sediment-associated contaminants by directly metabolizing them (Forbes et al., 2001).
Bioavailability is clearly neither a simple concept (Meyer, 2002) nor a simple parameter to assess. However, it is only the bioavailable fractions of contaminants that are of concern, hence the critical importance of determining bioavailability as part of the process of differentiating contamination from pollution and assessing not just hazard, but also risk.
Realistic exposure scenarios
Contaminants in sediments may reach organisms by several routes: from ingestion of sediment and associated pore water, from the pore water, from the overlying water. Different organisms will be exposed differently. For instance, burrowing organisms such as worms that ingest sediments and do not build burrows will not be exposed to overlying water, unless they move above the sediment surface. Organisms that build burrows will be insulated from the sediments and pore waters and will be exposed primarily via the overlying water. Similarly, epibenthic species on the sediment surface will be primarily exposed via overlying water.
The stability of sediments (and soils) when they contain contaminants that are or may become bioavailable is of paramount importance. The key question to be addressed is whether the sediment is liable to erosion resulting in exposure of deeper, more contaminated sediments of which a portion of the contaminants may be bioavailable, and/or causing bioavailable contamination down-stream (Haag et al., 2001; Chapman et al., 2002).
An excellent example of the importance of properly considering exposure routes concerns a recurring issue in aquatic sediments contaminated with PAH: the possibility of phototoxicity, specifically, increased toxicity as a result of photoactivated PAH accumulated in biota. In this case there is no question of bioavailability: the PAH are taken up by the organisms and are thus bioavailable. However, the vast majority of studies demonstrating PAH phototoxicity have been conducted in the laboratory under unrealistic exposure scenarios. Thus, although the phenomenon occurs, its significance in the natural environment remains to be established. McDonald and Chapman (2002) found only eight published studies that evaluated PAH phototoxicity in the field. However, none of these studies clearly demonstrated ecological relevance; all of these studies were flawed in some manner related to realistic exposure routes. Further, organisms have a variety of protective mechanisms (physiological, metabolic, behavioral) that protect against PAH phototoxicity, such that organisms which are exposed to sunlight have developed mechanisms to protect them against UV-light exposure and thus against PAH phototoxicity, while organisms that most commonly demonstrate PAH phototoxicity in the laboratory are not naturally exposed to sunlight (McDonald and Chapman, 2002). Proper consideration of exposure routes does not presently indicate that PAH phototoxicity is a pollution issue for sediment biota. There is a hazard, but no indication of risk.
Determining whether contaminants are bioavailable under realistic exposure scenarios is also particularly important because of the possibility of both direct (i.e., toxicity) and indirect effects resulting in shifts in ecosystem structure. Indirect effects may include effects at higher tropic levels due to changes in the number and diversity of intermediate trophic levels (Preston, 2002). This can result in simplified food webs and energetic bottlenecks in, for instance, metal-polluted lakes (Sherwood et al., 2002; Campbell et al., 2003) or altered decomposition processes in soil (Salimen et al., 2001). Indirect effects can also include the release of tolerant species from competition and/or predation. Both possibilities are illustrated in Figure 2. However, the ultimate effects of contaminants (or other stressors) on ecosystem structure may be ameliorated by functional redundancy, specifically more than one species having a similar role in ecosystem processes (Mermillod-Blondin et al., 2001).
Incorporating bioavailability and realistic exposure scenarios
Clearly, bioavailability and realistic exposure scenarios are essential components of any differentiation between contamination and pollution. Thus, Paracelsus' Dictum should be modified as follows: ‘All substances are poisons; there is none which is not a poison. The right dose of a bioavailable substance, administered under realistic exposure conditions, differentiates a poison. . .’.
Of course, modifying a Dictum is a lot easier than actually applying the additional components in, for instance, ecological risk assessment. In fact, our present state of knowledge regarding effects of contaminants in sediment (and other) systems is deficient. We cannot readily measure bioavailability on a routine basis. We cannot predict which are the realistic exposure scenarios for all species under all situations. And we cannot directly assess factors such as indirect toxicity.
However, we can use information gathered under the revised Dictum in a weight of evidence (WOE) framework to provide the best possible environmental information for decision-making (Burton et al., 2002). Weight of evidence frameworks that are logical, transparent, and readily understandable by lay personnel can serve to differentiate appropriately between hazard (the possibility of impact, Paracelsus' original Dictum) and risk (the probability of impact, the revised Dictum). Usually, the level of effort is proportional to the importance of the decision to be made. Such WOE frameworks can take the form of decision matrices, and are most useful if they are supported by (Chapman et al., 2002): a clear a priori rationale; detailed narrative explanation; and, subdivision generally into no more than three possibilities, high, moderate, low. These latter subdivisions match our present level of knowledge regarding environmental contaminant-related effects. We can generally differentiate areas that are not of concern (low) from those that are of extreme concern (high). We have problems differentiating intermediate areas (moderate).
While we do not have perfect knowledge, there is no excuse for not having the best information possible for decision-making. Modifying Paracelsus' Dictum (Figure 3) provides for the best possible information to differentiate pollution from contamination, and to move beyond hazard to risk.
This paper is based on a keynote talk given at the 5th International Symposium on Sediment Quality Assessment (SQA5), Chicago, IL, USA (October 16–18, 2002). I thank Drs. Allen Burton and Mohi Munawar for inviting me to give this talk and for paying my expenses.