Channel substrate prediction from GIS for habitat estimation in Lake Erie tributaries

The conservation of ecologically and economically important species, as well as the management of invasive species, benefits from the ability to make broad-scale predictions of habitat. In this paper, we revised an existing substrate size model based upon stream power to include variables that are readily-quantifiable in a Geographic Information System (GIS) (i.e. stream slope and drainage area). We found no significant difference between slopes measured in the field using surveying techniques and slopes measured in a GIS using a 10m digital elevation model and high resolution stream dataset. GIS-derived drainage areas and those measured with hand-delineations were also statistically similar. The revised model can be applied using both GIS and field-derived variables to predict median particle sizes from stream power in northeastern Ohio streams draining to Lake Erie. Integration of such models into a GIS could result in regional estimates of the amount and location of preferred fish habitat, which has important applications in fisheries management. In particular, we provide examples of how the predictive substrate model could improve assessment methodologies for invasive sea lamprey, thereby improving eradication measures, and how we may better understand geographic linkages between walleye spawning and nursery habitats.