The 1997 Student Survey
Our raw data was analyzed statistically using the computer program, SPSS. Our survey of 19 sites, totaling 76 m2 of the Cannon River, proved to be an insufficient sample size for most methods of statistical analysis. Proptera alata, the most abundant species found in our survey, comprised 42 of 61 total specimens. This prevalence suggests that the species has adapted well to the substrate and feeding conditions in DNR sites, IC and ID.
Our Proptera alata counts show that the species tends to group in locations of cobble sediment. In locations with cobble substrate, we found many more Proptera alata proportionally than we did in gravel or in sand beds. Our results are consistent with another mussel study (Downing and Downing, 1992) which showed the tendency for mussels to aggregate. Because site selection was necessarily random, we were not able to designate the substrate types surveyed. Sites surveyed were composed chiefly of cobble. It is therefore difficult to make statistical comparisons of species' substrate preference.
Diversity of samples
The diversity of species found in our survey was lower than expected. In the 76 m2 of the river searched, five different mussel species were found. The most recent Minnesota DNR mussel survey of the area (Davis, 1988) found thirteen different species, indicating greater diversity. This disparity suggests possible local extirpation since the last survey, but more evidence of this would need to be shown to make any affirmations.
We estimated there to be 122980 m3 of silt trapped above the Northfield dam. Removal of the dam would presumably release this mass of sediment, affecting fauna downstream.
In accordance with a study of sediment release in the Michigan's Pigeon River (Alexander 1986), we would expect the mussel populations below the dam to decrease. One factor contributing to this decline would be the reduction of certain host fish populations essential to mussels in their glochiodial stage. However, there have been no studies of fish- mussel species interactions in the Cannon. Therefore we can only speculate about the ramifications of fish host extinction. While we did explore the issue by communicating with other researchers who have studied fish host interactions, exacting such relationships was beyond the scope of this particular study.
Similarly, we would expect population declines in mussels as a result of the physical trauma of such a drastic increase in water turbidity. The silt build-up would presumably cause difficulty in mussel filter-feeding (Marking and Bills, 1979). Some mussels species, stuck under the mass of sediment, would be unable to emerge effectively (Marking and Bills, 1979).
Potential Sources of Error
We were unable to collect mussel specimens when heavy rains raised the river's water to dangerous levels. While attempting to collect data under these conditions, we lost control of the boat and both divers had difficulty staying near the dive boat. Some equipment was lost as well. Ideally we would have liked to have surveyed more sites to have a more representative sample of the river. However, weather conditions in July and August made it impossible to survey more sites using our methods.
Possible Future Studies
One way to build upon this project would be to further explore the relationship between mussel species and host fish in the Cannon River. If the dam were to be removed, the entire food chain of the river could be jeopardized by the release of sediment. Therefore, people planning dam removals, development, drainage etc. along the river should carefully analyze species interactions to see how they might be affected by human activities.
When considering the ramifications of removing the Northfield Dam, it is also important to understand how much silt would be released. Further geological studies of the sediment mass would help determine whether dredging the mud out of the river would be beneficial were the dam to be removed. Finally if planners wanted to dredge the area of sediment build-up, they would need a more exact figure of silt volume. While we arrived at a reasonable estimate using a mathematical model, our number was based upon several assumptions about the constancy of the river's shape and depth. Using seismographic equipment, geologists could get a more accurate estimate of silt build-up.
Another means of better ascertaining the status of mussels in the Cannon would be to standardize a search procedure and continue to survey the areas annually. Our results cannot be accurately compared with those of the Davis DNR survey (1988) because different protocols were used. The Department of Natural Resources, the Cannon River Watershed Partnership, and other concerned institutions could get a better idea of the overall health of the river if surveys were conducted in a regular manner.