Research prioritization

The research prioritization step is important in emphasizing research that is both impactful and achievable, and several methods are available to do this. Throughout the Nickel Tropical Research Program, research prioritization was evaluated and refined to ensure that the appropriate questions were addressed. Two potential methods that can be used for this purpose are a modified Leopold matrix and a sensitivity analysis approach. Both can be utilized to quantitatively evaluate the impact of additional data points on statistically derived toxicity estimates. Although these steps were not initially used in the research prioritization phase of the present program, they were utilized at later stages to ensure the prioritization of future research. In hindsight, these approaches could have been applied in phase 1 of the SEAM program and will be used in continued research efforts in these areas.

The Leopold matrix is a quantitative environmental impact assessment method pioneered in 1971 (Leopold et al. 1971). In the context of ERA, the Leopold matrix can be employed to evaluate and prioritize research needs in a specific region and identify gaps in knowledge for effects assessment of important habitat types and biogeochemical processes that control Ni bioavailability. A modified Leopold matrix based on uncertainty, importance, and practicality can be used to prioritize research that is both impactful and achievable.

The Leopold matrix approach relies on expert opinion to identify environmental compartments with gaps and to rank them accordingly on a scale of 1 to 5 for the following:

amount of certainty and data (focus on ecotoxicology) availability (1 = lowest and 5 = highest),

relative importance and impact (1 = lowest, 5 = highest), and

practicality (1 = hardest and 5 = easiest).

The sum of the 3 assessment scores can be used to rank priorities for research needs. This approach can be used at the beginning of the research phase to prioritize data collection in relation to the gaps identified in the prior step and could be repeated at the end of the research program to develop additional research needs or identify future research directions. An example of a Leopold matrix for future tropical risk assessment research is presented in the Supporting Information (SI2).

Additionally, a sensitivity analysis approach can be utilized to quantitatively evaluate the impact of additional data points on statistically derived toxicity estimates (e.g., hazardous concentration [HC] that adversely affects 5% or 50% of species, i.e., HC5 or HC50 values). The sensitivity of an existing species sensitivity distribution (SSD), and particularly thresholds derived from it such as the HC5 value, to additional ecotoxicity data could be assessed by the addition of hypothetical data to the data set and replotting the SSD (e.g., Wheeler et al. ²⁰⁰²; Wang et al. ²⁰⁰⁸). The hypothetical data used could be generated randomly, guided by the distribution of values already available, or could be selected arbitrarily. By replotting the updated SSD with a variety of additional data included, it would be possible to assess the potential impact of performing additional testing on an existing SSD. This approach would provide an indication of how many additional tests would be required, and the range of sensitivities that they covered, in order to affect the resulting HC5 value from the SSD by a predetermined amount. Taken together, the conceptual model, data gap analysis, and research prioritization provide a roadmap for Phase 2: Research and data collection.