Lack of Lateral Mixing Downriver of Tributary Confluences

Elaine Langone Center, Terrace Room

The river continuum concept describes changes to physical conditions, such as stream width, depth, and turbidity, that cause the habitat gradient of rivers to affect ecosystem metabolism. In studying metabolism throughout the Susquehanna River in Pennsylvania, we found evidence of lateral mixing patterns. We arranged our study sites into two groups based on major confluences: Group A - Milton (West Branch), Danville (North Branch), and their downriver site Isle of Que (Main Stem); Group B - Greenwood (Juniata River) and its downriver match Fort Hunter (Lower Stem). Numerical modeling of river water mixing calculated that lateral diffusion will be completed at a distance of 25 to 37 channel widths downflow of the confluence at low flow and high flow, respectively. Previous research shows faster lateral dispersion for stretches with sharp curves and more islands, so we expected relatively rapid mixing due to the many islands mid-channel throughout our study reach. If complete mixing were to occur 30 widths downriver, we would expect lateral diffusion to be complete approximately 25 km downriver of the group A confluence and 36 km downriver of the group B confluence. We did not observe complete mixing within these distances. To test extent of mixing, we took lateral transect samples at each site and compared upriver-downriver pairs: Milton-Que, Danville-Que, Greenwood-Fort Hunter, and Que-Fort Hunter. We found similar water chemistry for Milton-Que West, Danville-Que East, and Greenwood-Fort Hunter West. The Fort Hunter transect retained some similarity to the Isle of Que transect. Water chemistry similarities show a pattern between tributaries and corresponding sites downflow. Lateral stratification may have profound effects on river ecology, affecting predictions about species and their suitable habitats and having repercussions on ecosystem metabolism. Based on our observations, water sampling in large rivers should take multiple samples throughout a lateral transect: a single bank sample will unlikely be representative of the entire area. Further research on this may look to quantify exact river size (width, depth, flow rate) at which lateral incongruity becomes apparent or how tributary size affects lateral mixing.