Success or Failure? Do Indicator Selection and Reference Setting Influence River Rehabilitation Outcome?

Abstract

Recovery indicators play a crucial role in the evaluation of river rehabilitation projects. Various types of biological indicators are used to address different ecosystem attributes (structure, composition, and function) at different levels of the biological hierarchy (population, guild, and community). Indicator values are evaluated against reference information from various sources, representing the conditions to be achieved (near-natural references) or to avoid (degraded references). We studied the extent to which investigators’ conclusions on project outcome were influenced by the indicator and reference types used. We analyzed 40 selected studies dealing with the recovery of riverine fish assemblages after active rehabilitation of physical habitat and lateral connectivity. In 32 (80%) of the 40 studies, fish response was measured at the population level. Structural and compositional indicators dominated (31 and 24 studies, respectively), while functional indicators were underrepresented (5 studies). Eighteen studies used multiple indicator types for a given ecosystem attribute, a given hierarchical level, or both. Among these studies, we found only very limited evidence that project outcome differed among different indicator types (1 study). In contrast, highly heterogeneous results were found within the different indicator types at the level of the individual study (i.e., indicators addressing the same hierarchical level and ecosystem attribute resulted in different evaluations of project outcome). Such heterogeneity was related to the spatiotemporal variability of the results and species-specific responses to physical habitat rehabilitation. Most studies (73%; 29 studies) used a single type of reference, and the majority focused on degraded conditions. Among the 10 studies that applied multiple reference types, one-third (3 studies) showed inconsistent results (i.e., one reference comparison produced a positive assessment for a given indicator, whereas the second comparison indicated that the endpoint was not achieved). We discuss the implications of inconsistent project outcomes for future monitoring activities.

Received July 6, 2010; accepted March 7, 2011

ACKNOWLEDGMENTS

This study was part of the Rhone–Thur Project (www.rivermanagement.ch), funded by the Federal Office for the Environment; the Swiss Federal Institute of Aquatic Science and Technology; and the Swiss Federal Institute for Forest, Snow, and Landscape Research. Sharon Woolsey, Jan Heggenes, and two anonymous reviewers provided helpful comments on the manuscript.

Notes

^aLimited data availability due to a lack of information in the studies.