Bacterial contamination of surface waters is a common public health concern. Laboratory analyses can be expensive, yet such monitoring is important to ensure safe recreational opportunities and to produce effective watershed management plans. This project focused on the ability and preferences of volunteers using “home test kits” to evaluate their local streams and rivers using E. coli as an indicator of stream health. Six states (i.e., IA, IN, MI, MN, OH, and WI) participated in the study, with collaboration among University Extension, state regulatory agencies, and volunteers.

A training manual was produced and used in 35 training sessions involving 151 volunteers during the three year study. Since then, E. coli monitoring workshops have been conducted by project staff at various state, regional and national conferences, and the training and methods have been adopted in at least two other states. An on-line database was developed that holds all project research data and is now available for other states’ use. A journal article is currently being written for publication.

Volunteers were surveyed concerning their attitudes and perceptions of test kit usability. At the close of the study, two-thirds of volunteers preferred the 3M™ Petrifilm™.

Reliability results from this three-year investigation will provide crucial information for state agencies considering using volunteer monitoring data in watershed planning and perhaps TMDL development. Final conclusions included:

Of six test kits used in this study, four were found to be more reliable than the other two. The following conclusions pertain to Coliscan Easygel® incubated, 3M™ Petrifilm™, IDEXX Colisure®, and IDEXX Colilert®.

1. The test kits results compared fairly well with certified lab analysis. Regression analyses comparing test kits with certified lab analysis yielded R2 values ranging from .51 to .60 for all samples from all states combined.

2. 3M™ Petrifilm™ and the IDEXX methods were essentially equal in performance followed closely by Coliscan Easygel®, when data from states were combined, based on regression and analysis of covariance. Two out of three volunteers preferred 3M™ Petrifilm™ over Coliscan Easygel®. Coliscan Easygel® and the IDEXX methods are currently EPA-approved for water analysis, whereas 3M™ Petrifilm™ is not.

3. The test kits are good tools for screening and for justifying additional monitoring.

4. There was variability in the bacteria analyses among the certified labs used in the project. Due to natural variation in bacteria concentrations in surface waters as well as differences in holding times, the amount of variability within and between labs was not surprising.

Additional impacts of this research reach into each participating state and beyond. For example, several states have implemented credible data legislation which will place added importance upon volunteer data as state budget and agency resources continue to diminish. In Wisconsin, the Eau Claire County Health Department was interested in receiving split samples from project volunteers to analyze for E. coli bacteria. This provided the project an opportunity to assess inter-lab variability of results. The lab analyzed 21 samples for this project at no cost. In addition to being of assistance to the project, the data were then available to that county's health department, leading to increased information about water quality of local streams. In Minnesota, $50,000 in state lottery disbursement funds were received in 2006-2007 to extend volunteer bacteria monitoring and augment 406 grant funds. These funds would not have been secured without the 406 funds and the initial work of the project to identify reliable test kit alternatives. With these additional grant funds, MN volunteers were able to collect extra samples and calculate geometric means (5 samples within 30 days) to compare with state standards.

In Michigan a study is being conducted to assess if length of holding time of a water sample prior to incubation has an effect on measured E. coli counts. The study is a direct result of this project’s findings and recommendations that further work be done to determine potential variability in holding times.