Population Modeling: A Valued Instrument for Ecological Risk Assessments
In recent years, across industries, there has been a significant amount of effort invested in the application of population modeling for ecological risk assessments (ERAs). Waterborne’s increasing expertise in applied population modeling has contributed to the strategy and direction of ERAs through the development of case studies and the introduction of a systematic approach for the development of population models. Population models and other ecological modeling approaches can combine data and knowledge about species, its habitat, and exposure-effects relationships. Temporal and spatial aspects of a species’ habitat use and potential exposures can be considered explicitly.
A critical challenge to address is how to effectively implement population modeling within the ERA framework. Waterborne ecological modeler, Dr. Amelie Schmolke, and her co-authors have introduced a systematic approach for effective implementation of population modeling for pesticide risk assessments (Schmolke et al. 2017a). The proposed decision guide increases the efficiency and transparency of population model development, making population models more readily applicable in pesticide risk assessments. The decision guide is organized in four phases illustrated in Figure 1 for an example of herbaceous plants.
In Phase I, the model objectives are compiled systematically. The purpose of the model is defined in detail and aspects of the model may be determined prior to its development.
During Phase II, available data regarding the species of interest are compiled, as well as the pesticide exposure and the toxic effects relevant to the species. The resulting tables contain the information relevant for the model development along with the uncertainties in the data. Data gaps are identified systematically, and inform the details and assumptions in the conceptual model.
Phase III (decision steps) is comprised of the step-wise decisions needed to develop a minimal conceptual model. In this case, ‘minimal’ does not imply simplicity, but rather the “lowest level of complexity necessary to meet a given study objective” (Schmolke et al. 2017a). First, a life-history graph is prepared for the species of interest based on available data from Phase II. A series of seven decision steps are then followed by the model developer, addressing organism-level processes, temporal representation, spatial representation, density dependence, population status and environmental conditions, and indirect effects. The decision steps represent an iterative process with refinements to previous phases and the life-history graph throughout the minimal conceptual model development.
In Phase IV (Summary of the Minimal Conceptual Model), summaries of each decision step are compiled into the minimal conceptual model. Uncertainties in the data used for the model development and model assumptions applied are characterized during this phase. The summary also specifies which output metrics will be collected with the implemented model. The minimal conceptual model can be used to assess and adapt existing models for the current purpose, or it can be applied as a blueprint for implementation of a new model. Additionally, the minimal conceptual model may identify and prioritize gaps in the available data which may need to be filled before the implementation and application of the conceptual model.
Waterborne developed a conceptual population model of the Mead’s milkweed (Asclepias meadii), an herbaceous plant listed as threatened under the Endangered Species Act (ESA), as a case example for the decision guide (Schmolke et al. 2017a). The implementation of the conceptual model made use of a published population model of the species and adapted and extended it for use in pesticide risk assessment. Dr. Amelie Schmolke and her colleagues analyzed a range of scenarios representing exposure-effects relationships for two herbicides and their effects on the populations simulated with the model. Using the Mead’s milkweed population model, they were able to estimate population-level effects of herbicides over extended time periods, which exemplifies an ecologically relevant endpoint for ERAs.
With a population model for another threatened herbaceous plant species, Boltonia decurrens (Schmolke et al. 2017b), Waterborne estimated the potential population-level impacts of different herbicides on this short-lived species. In this case, conservative in-habitat exposure scenarios were combined with dose-response relationships for growth and survival of standard test species, based on standard vegetative vigor and seedling emergence tests, and applied to the species-specific model. Exposures were distributed across the simulated habitat applying the RegDISP model for spray drift, and a combination of the Pesticide Root Zone Model (PRZM) and the Vegetative Filter Strip Model (VFSMOD) for runoff. This distributed exposure modeling approach made it possible to assess potential effects of herbicides on plant populations growing in habitats that border chemical use areas and was applied to assess the effectiveness of spray buffer zones as mitigation measures.
In another population model approach, Waterborne addressed potential indirect effects of pesticides on fish populations. The listed slackwater darter (Etheostoma boschungi) was used as an example species. The darter species’ diet is comprised of aquatic insects and small crustaceans. Some pesticides could potentially affect the food availability of the species for limited time periods even if fish are not affected by the compound. With the model, such indirect effects to the simulated populations can be evaluated over extended time periods as well as assessment of different assumptions. Through a combination of species-specific life histories and direct and indirect effects, population models can play a significant role in determining the potential risks of a chemical to populations of listed and other non-target species.
Waterborne’s expanding depth and expertise in population modeling is continuing to provide a higher-level approach in ecological risk assessment. Contact Amelie Schmolke at firstname.lastname@example.org should you have questions or be interested in learning more about population modeling and how Waterborne is using it to support ecological risk assessment.