39-4 How Road Networks Affect Stream Fish Assemblages

Laura MacPherson , Renewable Resources, University of Alberta, Edmonton, AB, Canada
Michael G. Sullivan , Fish and Wildlife Division, Alberta Sustainable Resource Development, Edmonton, AB, Canada
A. Lee Foote , Renewable Resources, University of Alberta, Edmonton, AB, Canada
Cameron E. Stevens , Golder Associates, Edmonton, AB, Canada
Watercourse crossing structures are ubiquitous anthropogenic features across human modified landscapes. In the Rocky Mountain foothills of west-central Alberta, Canada, activities of the forestry and energy sectors have resulted in the construction of large supporting road networks and the installation of tens of thousands of stream-crossing structures. We performed physical and habitat assessments on 295 watercourse crossing sites in 15 different sub-basins in the summer and early fall of 2007, 2008 and 2009. Of the assessed watercourse crossing sites, we sampled for fish at 110 sites (32 bridge and 78 culverts crossing sites). We used bootstrapping analysis to examine how several culvert parameters (hang height, outlet plunge pool depth, water velocity, length and slope) altered the upstream abundance of eight fish species relative to reference bridge sites. As a secondary objective we determined whether there were differences in stream habitat characteristics upstream and downstream of bridge and culvert crossings. Physical drops at the outlet (hang heights), slope and outlet water velocities were the most important culvert parameters shaping upstream non-sportfish distributions. Culverts acted as complete barriers to burbot and partially impeded spoonhead sculpin, sucker and minnow movements. At culverts with high outlet water velocities (>0.5 m/s), abundances of burbot, spoonhead sculpin, minnows and suckers decreased by 0.32, 0.11, 0.34, 0.29 units respectively. Conversely, we did not find evidence that culverts acted as barriers to upstream passage of rainbow trout, rather, abundances of rainbow trout significantly increased upstream of the highest hanging, steepest and longest culverts. We suggest that culverts that exclude burbot, a voracious predator, provide a competitive release for rainbow trout upstream of culverts. Alternatively, higher temperatures above culverts may allow increased productivity of rainbow trout upstream of culverts. At inventoried stream crossings, water quality and substrate composition did not noticeably change upstream and downstream of bridges, while culverts had significantly higher water temperatures and more silt and sand upstream. Given the large number of culverts that may be acting as potential barriers in Alberta foothill watersheds, our research emphasizes the need to further our understanding of how stream fish species respond to characteristics of culverts and how watercourse crossings generally alter adjacent stream habitat in order to make informed regulatory and planning decisions.