MAIN CHANNELS, FLOWING BACKWATERS, AND ISOLATED BACKWATERS IN THE ST. CROIX (SACN) AND UPPER MISSISSIPPI RIVERS (MISS): NITROGEN CYCLING AT HIGH AND LOW FLOWS

William Richardson, USGS Upper Midwest Environmental Sciences Center, La Crosse WI
Lynn Bartsch, USGS Upper Midwest Environmental Sciences Center, La Crosse WI
Michelle Bartsch, USGS Upper Midwest Environmental Sciences Center, La Crosse WI
Richard Kiesling, USGS Minnesota Water Science Center, Mounds View MN
Brenda Moraska LaFrancois, National Park Service, Midwest Region, St. Croix Watershed Research Station

Connectivity among channels and flood plain habitats can exert considerable control over biogeochemical, water quality, and biologic processes in floodplain river ecosystems. Here we begin to characterize 1) spatial differences in nitrogen concentrations and biogeochemical cycling in main channels, flowing backwaters, and isolated backwater habitats; and 2) differences between high river flows and base flow in nitrogen concentrations and biogeochemical processes in the SACN and MISS.

The study included 2 sites, each, on the main channel (MC), flowing backwater (FBW) and isolated backwaters (IBW) on the St. Croix River (nr Close Slough at Osceola) and Upper Mississippi River near the 9th Street landing (St. Paul Park) and in River Lake (total of 12 sites among the two rivers). Water quality (temp., pH, DO, conductivity) was measured monthly (May - August), while nutrients (total nitrogen, nitrate-N [NO3], ammonium-N, total phosphorus, soluble reactive P) were measured in May, July, and August.

In May (spring flooding) and July (base flow) 2008 nitrogen biogeochemistry (sediment denitrification, nitrification and nitrate uptake rates, and sediment ammonia concentration) were measured at these sites in both rivers. Discharge in SACN during process measurements was 123 and 38 m³ s^-1 (May and July, respectively), and 338 and 65 m³ s^-1 in MISS (May and July, respectively). NO₃^- concentrations were highest in spring floods in MISS in all habitats during spring floods, decreasing as summer progressed (e.g., MC: 3.0 to 0.42 mg L^-1); but were lowest in SACN in spring floods and increased in all habitats as summer progresses (e.g., MC: 0.12 to 0.90 mg L^-1).

Highest rates of ambient denitrification (DEN, µg N cm-2 hr-1) were in FBW during May in both rivers (SACN: 1.2; MISS: 2.6) and lowest in IBW (~0.0) in both rivers on all dates. Potential denitrification (DEA, µg N cm-2 hr-1) was higher in MISS than SACN on all dates and in all habitats, and highest in the MISS in IBW in July (20.1). Highest rates of DEA in the SACN occurred in the IBW in May (7.5). With few exceptions, denitrification was NO3- limited in all sites in both rivers.

Nitrification (µg N cm^-2 hr^-1) was higher in MISS than SACN on all dates and habitats. Highest rates occurred in the MISS in IBW in July (5.6), whereas highest rates in the SACN occurred in IBW in May (1.3). Sediment NO₃^- uptake rate was strongly correlated with ambient denitrification (r²=0.53, P=0.013) indicating NO₃^- losses in many cases were due to denitrification under ambient conditions. Further, strong linkage between denitrification potential (DEA) and nitrification (r²=0.71, P=0.002) suggests the two processes were tightly coupled, particularly in the IBW, where delivery of surface water NO₃^- may be limited to periods of flooding. High rates of DEN in FBW were likely dependent on inputs from NO₃^-—rich spring (ground water)—a commonly observed feature of bluff-side channels.

Both "river" (SACN and MISS) and "habitat" (MC, FBW, and IBW) played a role in influencing rates of denitrification and nitrification—and in ways similar to other areas of the Upper Mississippi. Higher rates of N-cycling occurred in the MISS, perhaps related to generally higher concentrations of water column and sediment nitrogen. Isolated backwaters (IBW) in both rivers exhibited high potential for denitrification and nitrification. Seasonal patterns of NO3- concentrations likely reflect the character of upstream surface water carried during floods (N-rich in MISS; N-dilute in SACN), and perhaps late season dominance of groundwater inputs as surface water flows decline late in summer (esp. SACN).