Major rivers worldwide have experienced dramatic changes in flow, reducing their natural ability to adjust to and absorb disturbances. Given expected changes in global climate and water needs, this may create serious problems,including loss of native biodiversity and risks to ecosystems and humans from increased flooding or water shortages.Here, we project river discharge under different climate and water withdrawal scenarios and combine this with data on the impact of dams on large river basins to create global maps illustrating potential changes in dischargeand water stress for dam-impacted and free-flowing basins. The projections indicate that every populatedbasin in the world will experience changes in river discharge and many will experience water stress. The magnitude of these impacts is used to identify basins likely and almost certain to require proactive or reactive managementintervention. Our analysis indicates that the area in need of management action to mitigate the impacts ofclimate change is much greater for basins impacted by dams than for basins with free-flowing rivers. Nearly one billion people live in areas likely to require action and approximately 365 million people live in basins almost certainto require action. Proactive management efforts will minimize risks to ecosystems and people and may be lesscostly than reactive efforts taken only once problems have arisen.
Habitat unit classification can be a useful descriptive tool in hierarchical stream classification. However, a critical evaluation reveals that it is applied inappropriately when used to quantify aquatic habitat or channel morphology in an attempt to monitor the response of individual streams to human activities. First, due to the subjectivity of the measure, observer bias seriously compromises repeatability, precision, and transferability of the method. Second, important geomorphic and ecological changes in stream habitats are not always manifested as changes in habitat-unit or frequency or characteristics. Third, classification data are nominal, which can intrinsically limit their amenability to statistical analysis. Finally, using the frequency of specific habitat-unit types (e.g. pool/riffle ration or percent pool) as a response variable for stream monitoring commonly leads to the establishment of management thresholds or targets for habitat-unit types. This, in turn, encourages managers to focus on direct manipulation or replacement of habitat structures while neglecting long-term maintenance or re-establishment of habitat-forming biophysical processes. Stream habitat managers and scientists should only use habitat unit classification to descriptively stratify in-stream conditions. They should not use habitat unit classification as a means of quantifying and monitoring aquatic habitat and channel morphology. Monitoring must instead focus on direct, repeatable, cost-efficient, and quantitative measures of selected physical, chemical, and biological components and processes spanning several scales of resolution.