Applications of Causation Criteria to Ecological Science

Ecological risk assessors are often faced with problems of large scale and involving multiple impacted species and multiple stressors. The United States Environmental Protection Agency (USEPA) has devoted a substantial amount of effort to stressor identification including causation analysis derived from human epidemiology. Among other things, USEPA proposed a generalization of Koch's original postulates to include chemical toxicants:

• The injury, dysfunction, or other potential effect of the toxicant must be regularly associated with exposure to the toxicant in association with any contributory causal factors.

• The toxicant or a specific indicator of exposure must be found in the affected organisms.

• The effects must be seen when healthy organisms are exposed to the toxicant under controlled conditions.

• The toxicant or specific indicator of exposure must be found in the experimentally affected organisms.

USEPA also developed a general scheme for analyzing causation derived from the Bradford Hill postulates. In this scheme, USEPA proposes the use of four criteria that are based on associations derived primarily from the problem under consideration:

• Co-occurrence interpreted as the spatial co-location of the candidate cause and effect.

• Biological gradient often in the form of a dose-response relationship.

• Complete exposure pathway or the pathway that a stressor takes to reach a receptor.

In addition, USEPA proposes the use of five additional criteria that are based on either the problem under consideration or other situations:

• Consistency of association or the repeated observation of the effect and candidate cause in different times or places. The more observations and the more diverse the situations, the greater the case for causation.

• Experiment referring to manipulation of a hypothetical cause by removing the exposure rather than a laboratory test.

• Plausibility or the degree to which the cause and effect relationship could be expected in light of what is known about mechanism and stressor response.

• Predictive performance - whether the hypothetical cause has any initially unobserved properties that were predicted to occur.

Finally, USEPA proposes two criteria that are integrative in nature:

• Consistency of evidence - whether the hypothetical relationship is consistent with all lines of evidence.

• Coherence of evidence - whether a conceptual or mathematical model can explain any apparent inconsistencies among lines of evidence.

USEPA recommends considering all individual lines of evidence using a semi-quantitative scoring system and evaluating the total weight of evidence. USEPA also stresses the necessity for uncertainty and confidence analyses as components of the causation evaluation.

A paradigmatic ecotoxicologic problem is the North American Laurentian Great Lakes which have substantial ecological value as well as significance to Canada and the United States as fisheries and sources of drinking water. Over the past 50 years, environmental impacts to the Great Lakes have included population declines, reproductive failure, and deformities in fish, reptiles, and pisciverous birds, among other impacts. Multiple stressors associated with these problems include the presence of many persistent pollutants, discharges of nonpersistent toxic substances, overfishing, eutrophication, depredations of introduced exotic species, and degradation in habitat. For intervention in the form of regulation or remediation to be successful, it is desirable to be able to differentiate among the various stressors and determine which are causative agents associated with ecological impacts.

In 1991, Glen Fox was one of the first to propose the application of causation criteria to impacts in the Great Lakes. Fox recognized the parallel between human epidemiology and epizoology or ecoepidemiol-ogy which he took to be the study of ecological causes and effects in specific localities or among populations, communities, and ecosystems. Borrowing from patho-biology, Fox defined 'disease' to be any failure of normal homeostatic processes at any level of biological organization. With this broad definition in hand, Fox discussed the Bradford Hill and Susser criteria in the ecological context. For example, he recognized that criteria regarding specificity should be viewed in light of current biological thinking that many diseases have multiple causes and that a single substance may have a number of different biological effects with different underlying mechanisms.

The actual application of causation criteria to Great Lakes problems was undertaken by scientists from the International Joint Commission who evaluated associations related to chemical exposure and avian chick mortality and deformity and reproductive success of lake trout among other topics. In a paper published in 1997, Michael Gilbertson presented a case study applying epidemiologic causation criteria to the Great Lakes mortality, edema, and deformities syndrome (GLEMEDS). GLEMEDS is a syndrome observed in common terns and herring gulls in which there is a high incidence of chick mortality, edema, and deformity. No bacterial or viral pathogens were found on culture; however, eggs showed a high level of various persistent organochlorine compounds including polychlorinated dibenzo dioxins and furans (PCDD/PCDFs) and polychlorinated biphe-nyls (PCBs). The criterion of specificity was evaluated by the ability of specific PCDD/PCDF and PCB congeners to cause the syndrome in laboratory animals. The strength of the association was evaluated using correlations between PCDD/PCDF toxicity equivalents and egg mortality and deformity in nesting colonies of double-crested cormorants and Caspian terns. Time order was evident in the observation that bald eagles exhibited similar reproductive failure subsequent to the introduction of DDT in the 1940s in addition to other, more anecdotal, evidence. The observation that numerous species (including herring gulls, roseate, common, Caspian, and Forster's tern, and double-crested cormorants) showed similar effects was used to demonstrate consistency. Coherence was demonstrated by recourse to the scientific literature and determination that there was a biologically plausible mechanism of action. Gilbertson was able to conclude that the association between exposure to persistent organochlorine compounds and GLEMENDS was a causal one and recommended legislative and regulatory courses of action to mitigate the problem.

Another application of causation criteria is in the use of weight of evidence approaches for evaluating contaminated sediment. The causal hypothesis of chemical contamination impairment of benthic communities is typically evaluated using such an approach. The weight of evidence is developed from individual lines of evidence including:

• sediment contaminant chemistry and geochemistry,

• benthic invertebrate community structure as revealed by field observation,

• sediment toxicity testing, and

• consideration of bioaccumulation or biomagnification.

The first three of these lines of evidence constitute the well-known and widely used sediment triad. Wenning and co-workers have proposed seven criteria, derived from epidemiologic criteria, for using line of evidence data to evaluate causation in the context of contaminated sediment and impacted benthic communities including:

• co-occurrence or spatial correlation in which the impacted community co-occurs with the chemical contamination,

• temporal correlation,

• effect magnitude (corresponding to strength of association),

• consistency of the association at multiple sites,

• experimental confirmation in the field or in the laboratory,

• biological plausibility of the stressor-effect linkage, and

• specificity or consideration that the stressor causes a unique effect.

Adams and Collier conducted a test of the utility of causation criteria in a variety of ecotoxicological contexts. In this test, eight investigators were requested to use seven causal criteria to evaluate their individual studies. The criteria proposed included strength of association, consistency of association, specificity of association, time order/temporality, biological gradient, experimental evidence, and biological plausibility. Organisms involved in the individual case studies included fish, benthos, clams, and birds. Chemicals of potential concern included radionuclides, PCBs, poly-cyclic aromatic hydrocarbons (PAHs), silver, insecticides, metals, and complex mixtures. Toxicological endpoints included genetic and biochemical alterations, liver histo-pathology, reproduction, and population impacts. The utility of the criteria was judged on the basis of a semiquantitative scale similar to that used by USEPA using the following categories:

• convincing evidence presented that the criterion was met,

• strong evidence presented yet some questions raised,

• more likely than not or only little evidence presented,

• evidence presented for both accepting or rejecting the criterion or no available evidence for a particular criterion, and

• evidence presented that argued against accepting the criterion.

Application of this scoring system to the criteria for the case studies showed that specificity of association was the least useful of the criteria followed by temporality. The presence of multiple stressors was hypothesized as the basis of the lack of utility for specificity. This test also found that causation was more difficult to demonstrate at the community level of organization compare to lower levels of biological organization.

Solar Power

Solar Power

Start Saving On Your Electricity Bills Using The Power of the Sun And Other Natural Resources!

Get My Free Ebook

Post a comment