Compilation and Evaluation of Stream Restoration Projects: Learning from Past Projects to Improve Future Success

partially restored headwater stream in Pennsylvania


Over the past two decades, concerns about the degradation of Pennsylvania watersheds have led to a boom in the implementation of stream restoration projects. These projects are most commonly designed to improve instream and riparian habitat, stabilize streambanks, minimize the impact of non-point source pollution and enhance biodiversity. The State of Pennsylvania has been a leader in sponsoring these restoration initiatives through various programs supported by the Department of Environmental Protection, including Growing Greener and Stream Releaf. In addition, federal funds from sources including the United States Department of Agriculture have made it possible for large numbers of riparian buffer projects to be implemented through cooperative programs such as the Conservation Reserve Enhancement Program (CREP). The Pennsylvania CREP program has enrolled thousands of contracts that protect hundreds of thousands of acres in the Susquehanna, Potomac, and more recently the Ohio River basins (FSA fact sheet, 2004).

Given the growing investment in stream restoration efforts, there is an urgent need for tools to assess and improve the effectiveness of restoration efforts at local, state, and nationwide scales (Bernhardt et al. 2005). Increases in funding, education, and awareness relating to water quality issues have led to the growth of infrastructure facilitating the implementation of stream restoration projects. For example, watershed groups are increasingly prevalent and well organized. The private sector has expanded to fill consulting, project design, project construction, and permitting assistance roles. But what is the status of the documentation, evaluation, and scientific study of these projects? On the individual project level, are project objectives clearly defined and then achieved? Is the restoration community making the most of past practices by learning from them to increase future success? On a larger scale, are projects being completed as part of watershed-level implementation strategies to maximize benefits? This study and report represents a response to questions such as these.

The National River Restoration Science Synthesis

Our work compiling a stream restoration database for Pennsylvania and our direct involvement in a variety of stream restoration practices led to our participation in a working group known as The National River Restoration Science Synthesis (NRRSS, see task force brought together scientists and agencies from across the United States with the goal of compiling a National River restoration database. We collected data for Pennsylvania’s stream restoration projects in a manner that was consistent with NRRSS protocols and requirements, and project information we collected was included in this national database.

Executive Summary

Growing interest in and support for watershed protection and restoration in Pennsylvania has led to a marked increase in the implementation of stream restoration projects throughout the State. Although stream restoration has great potential to improve the health of watersheds, much remains to be learned about the best ways to maximize the environmental and social benefits of such practices. One of the primary motivations for this project was to increase such learning opportunities via an assessment of the environmental and social outcomes of restoration projects.

An initial goal of our project was to create a database of stream restoration projects in Pennsylvania. A comprehensive database permits the evaluation of project status and trends (e.g., the number of projects implemented, project locations, project implementers, project expenditures, frequency of different restoration practices). Such a database is essential for guiding the allocation of resources, designing monitoring strategies, and assessing project outcomes. We have assembled the most complete evaluation to date of stream restoration projects in Pennsylvania, with 1501 separate project records. This database is intended to be up to date and accurate for projects completed through 2002. Given the difficulty of obtaining information on all restoration activities in the state and continued implementation since we stopped compiling records, this is a minimum estimate of the total number of restoration projects. The number of projects per year increased from only 11 in 1996 to 451 in 2002. Like many other regions of the United States, water quality management projects with riparian restoration activities were the dominant type of stream restoration project in Pennsylvania.

In contrast to these major efforts to implement riparian restoration projects throughout the Commonwealth, much less attention has been focused on monitoring restoration projects to assess their ecological outcomes. A similar situation exists in other regions of the country. Without more and better monitoring, it is very difficult to evaluate how successful these projects are in meeting various watershed-level goals related to the improvement of stream health. Moreover, this lack of monitoring may compromise future funding opportunities if the effectiveness of current efforts is not validated.

We created a network of 70 riparian restoration projects throughout the state from which we gathered ecological data that can be used in future assessments of restoration outcomes. Instream and riparian conditions were monitored during 2004 at 30 restored reaches, each paired with a forested upstream reference reach. The paired structure of these data provides a particularly powerful design for assessing future ecological changes resulting from the restoration activity. This report also provides detailed information on our monitoring protocols to facilitate future assessments.

Our analysis of the differences between the ecological condition of restored sites and their paired reference reaches showed that the restored sites consistently scored lower in riparian habitat quality as well as the biotic integrity of both periphyton (i.e., attached algae) and benthic macroinvertebrate assemblages. These results clearly demonstrate that at the present time these stream reaches continue to exhibit the types of impaired conditions that originally made them candidates for restoration. This ability to readily detect differences between restored and reference reaches using established monitoring protocols and standard statistical procedures bodes well for the use of these methods in future assessments of restoration outcomes. This is relevant to this project and the evaluation of restoration efforts in general.

We used regression analysis to examine whether differences between the biotic integrity of restored and reference sites were related to differences in watershed and habitat properties between the paired sites. These analyses demonstrated that a number of macroinvertebrate response variables (e.g., pollution tolerance) were strongly related to several habitat variables (e.g., bank protection by vegetation). These results have two important implications. First, they support the view that restoration practices that shift riparian habitat conditions in impaired reaches towards those of forested reaches should lead to improvements in stream biotic integrity. Second, they suggest that one simple way to predict how stream health will respond to riparian restoration is by monitoring changes in riparian habitat conditions.

We also administered a social survey for 31 restoration projects to learn about the factors that influence project success, including design, implementation, maintenance, partnerships, outreach efforts, technical expertise, and public involvement. We learned that riparian restoration projects are often implemented to achieve aesthetic goals as well as environmental goals. Practitioners indicated that they commonly lacked sufficient funding to provide routine maintenance of restoration projects after they have been implemented. For example, 77% of projects required maintenance, but only 26% projects had funds dedicated for maintenance activities. Volunteer help is often recruited to prevent invasive plant species from overtaking native plantings, add additional plants, or repair fences. Similarly, practitioners recognize the importance of evaluating project outcomes, but they indicated that the project budget rarely included funds for monitoring. The result is that monitoring is either not carried out at all, or it is not conducted in a standard, organized way to maximize learning.

There are two critical areas where improvements are needed in stream restoration policies and practices. First, there is a need for a greater commitment to assess and learn from restoration outcomes. We propose a strategy for increasing the level of restoration monitoring to evaluate project outcomes, which will necessarily require that funders provide greater support and guidance for such monitoring. This increased commitment to project evaluation needs to be accompanied by the creation and long-term maintenance of a comprehensive and user-friendly database of restoration projects throughout the Commonwealth.


This report was prepared by Jamie Carr, David Hart, and Jim McNair.

Funding for this project was provided by the Pennsylvania Department of Environmental Protection’s Growing Greener Program and the William Penn Foundation. Support from these two funders helped us increase the scope and intensity of this study. Funds from William Penn made it possible to intensively study and catalogue restoration projects in Southeastern Pennsylvania. During the initial phases of the project we developed tools to focus field data collection and collect social survey information from practitioners. The Growing Greener funding made it possible to apply these tools to a study with a wider geographical scope that covered the entire state. The significant products of the statewide database and statewide project study from this project simply would not have been possible without both funding sources.

Many thanks are due to Patrick Center Scientists that contributed their knowledge and expertise to this project: Puneet Srivastava, Rich Horwitz, Tom Johnson, Becky Brown, and Don Charles.

We would like to thank all of the people that contributed their time, effort, and expertise to this project, including: the William Penn Foundation, Pennsylvania Growing Greener, members of the field crew that put in long days traveling and collecting data, members of Pennsylvania’s restoration community that were kind enough to take the time to speak with us and provide project and contact information, all of the members of the Technical Advisory Group meetings hosted at the Academy who shared their time and thoughts with us, members of the NRRSS working group, landowners, park directors, and organizations that granted us permission to sample on their property, the Keystone Stream Team, Melanie Wertz, Brian Vadino, John Hoekstra, Scott Cox, Joy Lawrence, Dan Salas, Jennifer Bennett, Ben Wright, Amanda Kindt, Angela Giuliano, and Mike Depew.

The views expressed here are those of the Authors and not necessarily those of the Pennsylvania Department of Environmental Protection.

Project Contact

For further information on this project please contact Roger Thomas at