Maintaining and restoring freshwater ecosystem services in the face of local and global change requires adaptive research that effectively engages stakeholders. However, there is a lack of understanding and consensus in the research community regarding where, when, and which stakeholders should be engaged and what kind of researcher should do the engaging (e.g., physical, ecological, or social scientists). This paper explores stakeholder engagement across a developing network of aquatic research sites in North and South America with wide ranging cultural norms, social values, resource management paradigms, and eco-physical conditions. With seven sites in six countries, we found different degrees of engagement were explained by differences in the interests of the stakeholders given the history and perceived urgency of water resource problems as well as differences in the capacities of the site teams to effectively engage given their expertise and resources. We categorized engagement activities and applied Hurlbert and Gupta's split ladder of participation to better understand site differences and distill lessons learned for planning comparative socio-hydrological research and systematic evaluations of the effectiveness of stakeholder engagement approaches. We recommend research networks practice deliberate engagement of stakeholders that adaptively accounts for variations and changes in local socio-hydrologic conditions. This, in turn, requires further efforts to foster the development of well-integrated research teams that attract and retain researchers from multiple social science disciplines and enable training on effective engagement strategies for diverse conditions.

In this work, we develop and test proxy-based diagnostic tools for comparing freshwater ecosystem services (FWES) risks across an international array of freshwater ecosystems. FWES threats are increasing rapidly under pressure from population, climate change, pollution, land use change, and other factors. We identified spatially explicit FWES threats estimates (referred to as threat benchmarks) and extracted watershed-specific values for an array of aquatic ecosystems in the Western Hemisphere (Ramsar sites). We compared these benchmark values to values extracted for sites associated with an international FWES threat investigation. The resulting benchmark threats appeared to provide a meaningful context for the diagnostic assessment of study site selection by revealing gaps in coverage of the underlying socio-environmental problem. In an effort to simplify the method, we tested regularly updated environmental and socioeconomic metrics as potential proxies for the benchmark threats using regression analysis. Three category proxies, aggregated from (i) external (global to regional, climate-related), (ii) internal (watershed management-related), and (iii) socioeconomic and governance related proxies produced strong relationships with water supply threat benchmarks, but only weak relationships with biodiversity-related and nutrient regulation benchmark threats. Our results demonstrate the utility of advancing global FWES status and threat benchmarks for organizing coordinated research efforts and prioritizing decisions with regard to international socio-environmental problems.