Many assessments will result in an alternatives matrix with multiple measures of services. These different measures make it difficult for analysts to consider changes in services or benefits on equal footing. If the goal is to combine results to produce one score or value for each alternative, only one method should be used to populate the matrix. Units must be reconciled to allow direct comparisons of options. If the units can be reconciled, an alternatives matrix can display the results for each option as well as an aggregation of the overall change in services and, perhaps, a dollar value for the expected change in benefits. Producing a valid aggregation (avoiding double counting, biases, and so on) is challenging and will require experts; therefore, in many cases it is an aspirational task. It can be accomplished in different ways given the methods used to populate the alternatives matrix.
When estimated changes in ecosystem services production are provided as ecological measures or benefit-relevant indicators (BRIs) (e.g., expected floods per year or numbers of desired bird species in viewing areas), they are often aggregated across services and options by measuring change relative to the status quo or baseline scenario (e.g., percent change). This approach does not make the units commensurate (fractions of flood risk are not comparable to fractions of bird species), and worse, it implicitly assigns a measure of relative importance to the services whereby the largest ecological change will drive the decision outcome even if it is not the most important change. Moreover, this approach reflects neither stakeholder preferences for the different levels of ecological change (small or big relative to the baseline) nor preferences for one service over another. This process can be improved by allowing decision makers and stakeholders to weigh in on each service separately and explicitly; this approach can be incorporated into monetary and nonmonetary valuation methods.
Monetary values for the estimated changes in ecosystem services production can be easily combined across services and options. It is important to ensure that the monetary values used are appropriate for the geography and context in which they are applied. Values for services are context dependent. It is also important to include all significantly affected services. If it is difficult to monetize all relevant services it will be important to find a way to fully include nonmonetized services in the decision process. One way to do this is to use nonmonetary (multicriteria decision analysis) methods that can combine monetary and nonmonetary methods.
In nonmonetary multicriteria decision analysis, methods are used to estimate stakeholder preferences for various levels of ecological performance (e.g., two or three more fish) as well as willingness to make tradeoffs among services (e.g., fish versus birds versus flood risk). These preferences can be combined with unitless measures of satisfaction for each service to obtain an overall measure of value for each alternative. These measures can then be combined across services and options to do a full comparison.
All aggregation methods have strengths and weaknesses that analysts should consider before and after their application to ensure that they are not biasing results or misrepresenting outcomes. Experts will be needed for this process.