Dos and don'ts of spatially explicit ecological risk assessments

Abstract

Location affects exposure and response to stressors at scales ranging from small sites to large regions. Tools such as geographic information systems (GIS) make it feasible to conduct spatially explicit ecological risk assessments (ERAs). However, no tools provide a panacea, and complex models based on sparse data can be inappropriate and misleading. Operations such as an interpolation within a GIS are models, and so they contain assumptions and uncertainties. Errors can propagate easily as numerous operations are performed, and the resulting uncertainties should be made explicit. Analysts should assure that space actually matters by using explanatory data prior to investigating spatial correlation. Exposure and risk estimates from experimental plots or monitoring stations may require scaling factors if they are to be used for larger watersheds or regional analyses. Maps are useful because they present complex spatial information in a manner that is easily interpreted. However, they must be prepared and interpreted with caution because they can suggest that there is a great deal more information than actually exists. In addition to presenting maps of predicted risks, maps of the spatial patterns in the uncertainties in these risk estimates should be included in risk assessments.

Relevance of Spatial and Temporal Scales to Ecological Risk Assessment

“…For example, responses of plants in the laboratory or small-scale field plots to ozone may be used to model changes in regional forest productivity (Laurence et al 2000). This approach requires extrapolation of effects across scales, which is difficult and risky (Woodbury 2003), but is often demanded to generate policy. Policies are a product of human institutions and operate at particular temporal and spatial scales.…”

Identifying Modeled Ship Noise Hotspots for Marine Mammals of Canada's Pacific Region

“…We say “may be”, because indirect effects of noise can be mediated by effects on prey or predators, which is beyond the scope of this study. We illustrate some key elements of a spatially explicit risk assessment [35] for marine mammals and noise in the northeast Pacific, by integrating information on average distribution and abundance of 10 marine mammal populations, cumulative acoustic energy from ships, and our best estimate of how that acoustic energy may be perceived by the auditory system of the various marine mammals in question.…”

Evaluating Critical Uncertainty Thresholds in a Spatial Model of Forest Pest Invasion Risk

“…Recently, there has been much literature discussing the characterization of uncertainty in general and ecological risk analysis (14,17,20,21,24,25) as well as in specific invasive alien species contexts. (5,26,27) Similarly, Li and Wu (22) provided an overview of un-certainty analysis methods for large-scale ecological models used as environmental decision support.…”

“…(9)(10)(11) As a result, the historical emphasis in invasive alien species risk management has been on preventing introductions rather than on the development of comprehensive response strategies for all potentially impacted geographic areas; thus, spatial aspects of risk have often been neglected, (12,13) even though total risk of both invasion and impact clearly varies between geographic domains. (14) Nevertheless, risk maps, as the spatially explicit realizations of risk assessments, have become increasingly popular among decisionmakers and regulators as support tools to allocate resources for quarantine, monitoring, and control of invasive alien species. (9,15)…”