The project began in 1992 and is intended to continue indefinitely. Study site selection is influenced both by the goal of a complete study of the river system and by the interest of volunteers in a local site. New sites are added on an irregular basis, as volunteer and personal or community interest occurs. Benthic populations are monitored for three years to get an idea of the variation in the population, after which the sites are monitored between 1 and 3 times per year. Sites are monitored more frequently when the population appears to be changing.

DATA QUALITY OBJECTIVES AND METHODS FOR MEASURING THE BENTHIC POPULATIONS
The primary requirements of our benthic population study are thorough sampling and reliable comparisons both between sites and at each site over time.

Macroinvertebrate Sampling Strategy: Since our evaluation is based on the diversity in the population, we attempt to include a complete sample of the different groups present, rather than a random sub-sample. We do not assume that a single collection represents all the diversity in the population but rather we consider our results reliable only after repeated collections spanning at least three years. Our results are compared with other locations in the same river system that have been sampled in the same way. Most sites are sampled by different collectors at different times to diminish the effects of bias in individual collecting styles. Samples where the population looks different than past samples at this site are usually double-checked by a new team within two weeks.

Sampling Process: Multiple collections are taken from each habitat type present at the site, including riffle, rocks or other large objects, leaf packs, submerged vegetation or roots, and depositional areas. The streamside leader records the locations on a map. The collector transfers the material from the net into white pans. The volunteers pick out samples of all different types of macroinvertebrates from the pans and place them into jars of 70% ethyl alcohol. During the collection, the collector provides information in response to questions on the data sheet that reviews all habitats to be sampled, the state of the creek, and any unusual steps or observations. The streamside leader assists other team members in finding small creatures in the pan.

Sample Handling and Custody Requirements: At the collecting site, all jars receive a label written in pencil, starting date, location, name of collector, and number of jars containing the collection from this site, which is placed inside the jar. The data sheet also states the number of jars containing the collection from this site. The team leader is responsible for labeling and securely closing the jars, and the team Manager is responsible for returning all jars and all equipment. Upon return to the Watershed Council building, the collections are checked for labels, the data sheets are checked for completeness and for correct information on the number of jars containing the collection from the site, and the jars are secured together with a rubber band and site label until they are examined and counted on the day of identification. The data sheets are used on the identification day, after which they remain on file indefinitely. At the time of identifying the sample, the data sheet and the jars are checked to ensure that all the jars, and only the jars, from that collection are present prior to emptying them into a white pan for sorting. If any creatures are separated from the pan during identification, a site label accompanies them. When identification of a sample is complete, the entire collection is placed in a single jar of fresh alcohol with a polyseal cap and a printed label inside the jar and stored indefinitely.

Comparability (how well our results could be compared to other similar studies):
We intend that our results be thorough and reliable. Toward that end we seek opportunities to improve our performance. Our protocol was designed by professional aquatic ecologists and is similar to the collecting protocol established by the State biologists (MDEQ). The primary differences between our protocol and that of the State are: [1] we do not select a subset of creatures randomly from the collected sample; [2] we often collect somewhat more than 100 creatures; and [3] we do not identify our collection at the study site. While our population sample is sometimes affected by the bias of the people who are collecting creatures out of the sample pans, we do multiple collections (over three years) to assess a site, allowing a variety of people to add to the information. We conduct a very reliable identification of the collected sample in a laboratory setting with aquatic entomologists doing the identifications, and should any questions arise, we can reexamine the stored collections.

Our performance was compared with that of an aquatic entomologist, by the entomologist (C. Riseng, Fluvial Ecosystems paper, 1995). Five of our sites in two streams (one urban and one rural) were thoroughly sampled by her for one hour each, one month after our collection, to compare the thoroughness of our results with hers. She concluded that the collections were mostly similar, especially for the parameters that we use in site analysis. She did, however, find a different number of Families at each site. She compared her collections to an average of our three years of data. She found two and four more families than we did (each in two locations), while in one location we found five more families than did she. A Paired T-test found no statistically significant bias in our sampling. The Table below shows the number of taxa found in each sample in 1995, and also the total number of taxa in the composite of several samples collected in the years shown. The individual collections were only 1/3 to 1/2 of the total population, except for the site that has an unusually small population (Scheffler).

The comparability of our performance was measured against the DEQ site evaluations on five sites in 1997, as shown in the table below. Different teams of our volunteers collected from the same sites as the biologists from the DEQ within two to five days of each other, four in June and one in September.

Using the figures reported by the DEQ, the number of taxa found by each team was similar in three comparisons (the greatest difference was 27 vs 23), while the DEQ found much greater diversity at two locations (31 vs 21, and 22 vs 9). There was no systematic difference between our two sample sets. The number of sensitive families was the same at nearly all sites, while the number of EPT families differed by as many as four (10 vs 6). Usually the DEQ found more EPT, but at one location the volunteers found twice as many caddisfly families as the DEQ. While some teams found more variety than others, our volunteers did a good job of sampling the population compared to the professionals.

The number of taxa reported by the DEQ did not match the number in their collection jars. In the four jars we were given, three contained about half as many taxa as reported from the field identifications, and one had only one-fifth. The taxa that were reported but absent from the jars included many families of mayflies, caddisflies, beetles, some diptera, hemiptera, and odonata, along with molluscs, crutacea, leeches, flatworms, and bryozoa. The only EPT family consistently found by the DEQ and not by the Volunteers was Perlodidae. In those DEQ collections that we have examined, the Perlodidae were mis-identified; they are Perlidae. The only taxa that the Volunteers did not find was Bryozoa. Taxa found at some sites by the DEQ but not by the Volunteers included some families of hemiptera, molluscs (especially snails), amphipods, isopods, and flatworms. (Please note that these groups are commonly found by the Volunteers, who have also found Bryozoa.)

We shall continue to learn from the professionals whenever we can arrange the opportunity. In fact, Mike Alexander, aquatic biologist with the DEQ, attended our Volunteer training in March, 1998. He appraised the training as follows, "Your training is very thorough. It is good the way you have those with more experience paired up with the less experienced. I have no improvements to suggest. You have a good system, with many checks and balances. It seems very solid."

METHODS FOR ADDITIONAL MONITORING METRICS
Methods for Measuring Habitat Quality: Habitat assessments must be completed on a single day. A descriptive procedure guides the team to make observations and map a 300-foot stretch of the creek or river and take detailed measurements over half of that area, or 150 feet. Pictures are used to document areas of erosion and anything that the team is unsure how to characterize. When it is difficult to pick a single number, for example 60% of the banks were bare, the team is encouraged to write down a range of numbers such as 50-75% of the banks were bare. The habitat assessment includes a mixture of objective and subjective measurements, with some repetition, to assure accuracy of the information. When discrepancies occur, the team is contacted as soon as possible and issues are resolved through discussion or another steward visits the site to verify the information.

Macroinvertebrate Sampling and Habitat Monitoring Process Design: Monitoring sites are identified by the name of the county, the creek and the road crossing or other distinctive landmark. Road maps and field-drawn maps showing the location of each site are provided to the collecting team. Extreme conditions of weather or stream volume are noted on the field data sheet. Permission to access property is obtained at least one week prior to the designated date.

Methods for Measuring the Conductivity at Each Site: Samples of stream water are collected in plastic jars with plastic caps right after the jar has been rinsed three times with water from the stream. Conductivity is measured in all the collection jars at one time at the end of the sampling day, with a Myron L Tech Pro meter that has been calibrated with a potassium chloride standard for 1413 uS per cm. After discarding the samples, the jars are rinsed three times with tap water and allowed to air dry before storage.

Methods for Measuring Temperature Extremes at Each Site: Volunteer stewards secure a max/min thermometer underwater at a location of moderate depth and out of direct sunlight. They record, on a weekend day, the maximum and minimum temperatures reached during each week of July and August. After each reading is taken the thermometer is reset. Volunteers used mercury (Taylor #5458) or digital (DeltaTRAK) max/min thermometers.

Training Requirements: The volunteer group leaders are trained for each activity and lead the teams of volunteers when monitoring the macroinvertebrate populations or measuring the habitat quality of a site.

Collectors attend a 4-hour training session in the creek to learn and practice the methods of collecting samples from all habitat types. Training sessions are offered a few weeks prior to the spring and fall collections. In addition, collectors review the methods in the morning just prior to collecting, and the Field Data Sheet includes sampling categories and questions that guide the collector when they are in the stream.

Team leaders assessing habitat quality attend a 3-hour training. At the Training volunteers learn how to "read a river" by examining characteristics of the stream such as the stream banks, measuring the stream widths and depths, and scrutinizing what type of material (such as sand and gravel) is on the stream bottom. Stewards also learn how to make a simple map of the locations of various features, such as pools and riffles that are important homes for aquatic animals. The training begins with some illustrated classroom instruction, then the group carpools to a nearby creek to practice the measurements and discuss some of the more subjective measures of habitat quality.

Records: The "streamside leader" is a trained volunteer, familiar with the procedures and datasheets, who leads each monitoring team. During the benthic collections, streamside leaders record the following information at each site: [1] number and locations sampled in each habitat type, [2] answers to questions about specific techniques and about deviations from protocol.

All macroinvertebrate collections are stored in alcohol and kept indefinitely as vouchers. The datasheets include the name of the biologist who identified the collection. Teams assessing the habitat quality are guided by a list of procedures and an informative datasheet. In addition to keeping all of the original paper datasheets, the information is stored, managed and analyzed in ACCESS.

Sampling Methods Requirements: Equipment for benthic macroinvertebrate samples includes a D-frame collecting net (mesh size 20 x 24 mesh/inch) and featherweight forceps. Transfer from the net into white pans is facilitated by use of large squirt bottles to rinse out the debris and the macroinvertebrates. Prior to leaving each site, the net is thoroughly rinsed and examined to ensure that no creatures are carried in it to the next site. Sites that have zebra mussels are the last site visited.

Equipment for measuring the habitat quality includes a tape measure and measuring stick marked in feet and tenths of feet to avoid confusion about unit measurements.

Analytical Methods Requirements: Benthic population and habitat assessment methods are based on the state’s Procedure #51. However, rather than collect a random sample of macroinvertebrates, we attempt to collect all the different kinds of macroinvertebrates to measure diversity. Since we do not attempt to have a quantitative sample, randomness and a uniform sampling effort are not required. This allows less experienced collecting teams to take additional time in order to be thorough. The length of time spent at each site is noted on the data sheet and is usually between 45 minutes to 1 1/4 hours. Two collectors sample large sites (over 100 feet wide) on the mainstem of the Huron River.

We have added many objective measurements to the habitat assessments. For instance, the team examines ten transects along which ten or more measurements per transect are made. The measurements include stream depth and the size of the substrate at that point. At the same time, the width of the stream and the channel, the depth of the channel, and the angle of the banks are also recorded. We are particularly interested in the composition of the substrate and believe that, with well over 100 measurements, we get a reasonable idea of they substrate composition at that time.

The population sample is identified to family level by trained aquatic entomologists; the name of the entomologist is recorded on the data sheet. Any unusual samples or insects that are difficult to identify are double checked for accuracy a few days after the identification day. When volunteer monitors have learned to identify the Families, their identification is checked by one of the entomologists. Volunteers and entomologists work together on the samples on one day, usually two weeks after the collection period. The responsibility of the volunteers is to separate the collection into look-alike groups (families), count the numbers in each family after it is identified by the entomologist, and record the numbers on the data sheet. The families are placed in separate compartments of a tray and labeled with a letter that corresponds to the letter written next to the family name on the “Count List” data sheet.

Data Management: Field data sheets are checked by Program staff for completeness upon return to the Watershed Council office; any omissions or confusions are clarified as soon as possible. All data is entered into Excel spreadsheets designed for the project or the Access database which are both useful for analysis as well as for production of stream reports. The final tables of data are checked for accuracy against the raw data sheets after every time of entry.

Last Updated: February 2003