Evaluating multiple lines of evidence requires a
weight-of-evidence assessment, which may be qualitative, semi-quantitative or quantitative (each becoming less reliant on best professional judgement). Ideally, an assessment should involve minimal best professional judgement (Dafforn et al. 2016) so independent assessors of the lines of evidence will reach the same conclusions.
Be mindful that an overly complex assessment may be costly and require larger, often impractical, data inputs.
Our approach to weight-of-evidence assessment mainly focuses on the evaluation within a single pressure, acknowledging the need to acquire similar information across similar and relevant lines of evidence for any additional pressures.
Human activities inevitably result in multiple pressures and stressors on aquatic ecosystems. The modern challenge lies in developing cause-and-effect relationships and diagnostics in complex ecosystem–pressure–stressor scenarios (Baird et al. 2016, Pistocchi et al. 2016).
A thorough review of methods for assessing multiple pressures and stressors on aquatic ecosystems is beyond the scope of the Water Quality Guidelines but guidance will be required in the future.
Dealing with the complexity amongst different pressures and stressors to draw correct inference is partially acknowledged and addressed elsewhere in the Water Quality Guidelines:
- For waters (including effluents) or sediments representing a complex chemical mixture
- Some traditional or improved
approaches to isolate the effects of multiple stressors on ecosystems.
Criteria-guided judgment based on human epidemiological precepts (Hill 1965) may be useful for inferring cause in complex weight-of-evidence evaluations.
- For evaluating the combined effects of multiple pressures and stressors,
semi-quantitative methods have been used. Evaluating the specific causal links from lines of evidence to particular component pressures will rely on a good conceptual model and an informed system understanding. In many instances, only one pressure will dominate so the cause, and thus the focus of management action, will be obvious.
Baird et al. (2016) introduced seminal papers arising from a workshop on multiple-stressor impacts on aquatic ecosystems. Papers by Chariton et al. (2016) and Dafforn et al. (2016) from that workshop reviewed current approaches to evaluating multiple pressures and stressors. They also addressed the development of improved diagnostics in complex ecosystem–pressure–stressor scenarios, including genomics, modelling and statistical tools.
Evaluation methods for multiple pressures and stressors
In its simplest form, evaluation of multiple pressures might integrate the findings of independent assessments of the ecosystem effects associated with particular pressures into a scientific consensus statement. Such a statement might contain management recommendations agreed to by proponents of each single pressure assessment.
An example is the scientific consensus reached with respect to land use effects on the water quality of the Great Barrier Reef (Waterhouse et al. 2017).
Cumulative impacts mapping
Cumulative impact mapping (or cumulative effects mapping) uses spatial data of the pressures and stressors in an ecosystem to build in some measures of the ecosystem’s sensitivity to those pressures and stressors.
Early applications of cumulative impact mapping assessed the effects of human activities on the world’s oceans (Halpern et al. 2008). A standardised method based on expert professional judgement is used to weight a range of drivers (pressures) that are then combined into single cumulative impacts for each ecoregion. Drivers were quite diverse and included such things as commercial shipping, commercial fishing, species invasion and benthic structures. For more localised studies of coastal Californian ecosystems, additional drivers included coastal power plants, marine debris, invasive species and oil rigs (Halpern et al. 2007).
Statistical classification methods
The European Union used statistical classification methods, such as
regression trees, logistic regressions and
random forests, as it worked to improve the ecological condition of rivers in Europe. Such methods enabled the contribution of different pressures (contaminants, hydrological alterations, geomorphological alterations, land use) to be attributed to the ecological status of rivers (Pistocchi et al. 2016).
Weight-of-evidence evaluation methods using Bayesian network calculations
The use of Bayesian methods for assessing causality has been discussed in several publications (Carriger et al. 2016, Linkov et al. 2015). Bayesian networks can enhance weight-of-evidence assessments by probabilistically examining the strength of associations between different lines of evidence. While regarded as a significant advance over qualitative and semi-quantitative approaches, the lack of feedback loops and the complete lack of development of posterior distributions in
many implementations of Bayesian networks have been subject to criticism.
Baird DJ, Van den Brink PJ, Chariton AA, Dafforn KA & Johnston EL 2015,
New diagnostics for multiply stressed marine and freshwater ecosystems: integrating models, ecoinformatics and big data, Marine and Freshwater Research 67(4): 341–392.
Carriger JF, Barron MG, Newman MC. 2016,
Bayesian networks improve causal environmental assessments for evidence-based policy, Environmental Science & Technology 50(24): 13195–13205
Chariton A, Sun M, Gibson J, Webb A, Leung K, Hickey C & Hose G 2016,
Emergent technologies and analytical approaches for understanding the effects of multiple stressors in aquatic environments, Marine and Freshwater Research 67: 414–428.
Dafforn KA, Johnston EA, Ferguson A, Humphrey C, Monk W, Nichols S, Simpson S, Tulbure M & Baird DJ 2016,
Big data opportunities for assessing multiple stressors across scales in aquatic ecosystems, Marine and Freshwater Research 67: 393–413.
Halpern BS, Walbridge S, Selkoe KA, Kappel CV, Micheli F, D’Agrosa C, Bruno JF, Casey KS, Ebert C, Fox HE, Fujita R, Heinemann D, Lenihan HS, Madin EMP, Perry MT, Selig ER, Spalding M, Steneck R & Watson R 2008,
A global map of human impact on marine ecosystems, Science 319: 948–952.
Halpern BS, Selkoe KA, Micheli F, Kappel CV 2007,
Evaluating and ranking the vulnerability of global marine ecosystems to anthropogenic threats, Conservation Biology 21: 1301–15.
Hill AB 1965,
The environment and disease: association or causation? Proceedings of the Royal Society of Medicine 58: 295–300.
Linkov I, Massey O, Kiesler J, Rusyn I & Hartung T 2015,
From weight of evidence to quantitative data integration using multicriteria decision analysis and Bayesian methods, ALTEX 32: 3–8.
Pistocchi A, Udias A, Grizzetti B, Gelati E, Koundouri P, Ludwig R, Papandreou A and Souliotis I 2016,
An integrated assessment framework for the analysis of multiple pressures in aquatic ecosystems and the appraisal of management options, Science of The Total Environment 575: 1477–1488.
Waterhouse J, Schaffelke B, Bartley R, Eberhard R, Brodie J, Star M, Thorburn P, Rolfe J, Ronan M, Taylor B & Kroon F 2017,
2017 Scientific Consensus Statement: Land use impacts on Great Barrier Reef water quality and ecosystem condition, State of Queensland, Brisbane.