USING GENETIC ALGORITHMS TO EXPLORE THE EFFECTS OF CHANGES IN HYDROLOGIC VARIABLES AND SUSPENDED SEDIMENT ON FISH BIODIVERSITY, SANDUSKY RIVER, OHIO

Large amounts of research have been conducted on the importance of hydrologic variables and suspended sediment of streams on fish habitat. The typical hydrologic variables used in these studies include parameters such as peak, low, and average flows. Much of the suspended sediment research has been based on the concept of a lethal threshold, which does not account for fish leaving an area. The Sandusky River in Ohio is an important river both as a fish nursery for Lake Erie and for recreational and commercial fishing in itself. This study attempts to answer two questions on the Sandusky River. First, how does varying flows and suspended sediment (mostly due to land use changes) modify fish diversity? Second, is a genetic algorithm capable of determining relationships between multiple dependent and independent variables in a large dataset?

Data from the USGS gage in the Sandusky River as well as fish data collected by the Ohio Environmental Protection Agency and Ohio Division of Wildlife during 16 years was used in this study. The Index of Hydrologic Alteration program (IHA) was used to generate different hydrologic and suspended sediment variables that were then run through a genetic algorithm to determine different biodiversity indices such as the Shannon Index, the Simpson Index, and the Species Richness.

The genetic algorithm produced a single equation with all the variables deemed important to a specific biodiversity index using multiple dependent and independent variables. Both flow and suspended sediment were shown to be important at multiple levels in the Sandusky River. Common variables were found among all three biodiversity indices. These results also show the many complex factors of flow and suspended sediment that need to be maintained in order to obtain optimal biodiversity in the Sandusky River.