Nutrient Fluxes and Removal Mechanisms in a Controlled Drainage and a Pond-Wetland System Treating Cropping System Drainage in Eastern Ontario, Canada

Abstract

Tile drainage, a necessity in humid continental cropping systems, acts as a direct conduit for nutrient- and sediment- rich water to surface water bodies. With increasing agricultural intensification, this pathway has become a significant non-point source (NPS) of pollution, contributing to anthropogenic eutrophication and downstream hypoxia. Environmental directives targeting agricultural NPS pollution, such as the Canada-Ontario Lake Erie Action Plan and Ontario’s Domestic Action Plan, emphasize targeted, on-farm Beneficial Management Practises (BMPs) that deliver multi-pollutant mitigation and are supported by practical, evidence-based knowledge sharing across jurisdictions. The central objective of this multi-year, full-scale research study was to evaluate two BMP technologies - controlled drainage (CD) and pond-wetland systems - for treating cropping system drainage during non-frozen periods (April to November) in cold-climate regions, such as Eastern Ontario, while addressing technology-specific knowledge gaps related to their performance and removal mechanisms. While CD studies in these regions often focus on the growing season, the fate of retained water and nutrients after stoplog removal (flush) and potential CD residual effects during post-harvest period remain unclear. In this study, CD optimally maintained at 0.38–0.44 m below the soil surface during the Growing Period (GP) reduced flow and nutrient loads by 51- 76% compared with free drainage (FD). Stoplog damming significantly reduced solids (TSS) and particulate phosphorus (PP) export, created conducive environment for nitrate reduction during dry GP periods, and potentially enhanced sorption of soluble reactive phosphorus (SRP) - increasing water and nutrient availability during crop-stress periods. During the Flush Period, flow and nutrient losses increased at CD before returning to background levels. Evidence of denitrification was indicated by nitrate gradients in the dammed soil profile during flush and a 35% higher relative abundance of nitrate-reducing bacteria under CD compared with FD. No residual GP effects were observed in post-harvest period, except for SRP. Overall, CD implemented in GP effectively offset the hydrological and nutrient exports of the Flush and Post-Harvest Periods, reducing total flow (37%), nitrate (21%), and SRP (57%) load from planting to freeze-up. Although warmer-climate pond and constructed wetland systems report high nutrient removal, their long-term effectiveness in cold-climate, treating cropping systems remains uncertain, as attenuation of nitrate-rich, low-carbon drainage can become temperature-limited. In this 5-year study, 0.4 ha pond-wetland system (system-to-catchment area ratio = 0.93%) and varying mean depths of 0.1 to 1.3 m, exhibited strong seasonal and annual variations driven by flow and internal processes. Overall, inlet-outlet concentration monitoring, revealed that the system was most effective in retaining TSS (49%), followed by nitrate (30%), PP (22%), and SRP (14%), suggesting system functioned as a net sink for all parameters. However, seasonal patterns showed high variability, from intermittent to negative removals in colder, high-flow periods (Late Spring, Fall and Spring-melt) to consistent highest removals in warmer, low-flow Summer on both concentration and mass basis. However, high relative removal during low-flow periods did not necessarily translate into substantial reductions in annual exports. Internal monitoring revealed pronounced pond-to-pond gradients, uniquely providing insights relative to different depths. The deepest, unvegetated pond (mean depth = 1.3 m) achieved the highest retention of TSS and PP via sedimentation. In contrast, the shallow inlet pond (0.1 m) consistently promoted resuspension and algal growth, acting as sources of solids and PP, especially during Summer and Fall. The terminal, wetland-like outlet pond (0.3 m) provided no additional treatment benefits across all parameters under the observed hydraulic and loading conditions. SRP dynamics was highly variable, showing no consistent system physical controls, indicating internal biological and geochemical processes can exceed net retention in this low P loading system. Denitrification kinetics were best described by a cascading pond-to-pond areal-based first order P-k-C* model (k = 0.14 m/d; theta = 1.19), outperforming traditional inlet-outlet and volumetric (HRT-based) formulations, indicating nitrate removal scaled more strongly with benthic surface area than with water-column volume. Consistent with this, kinetic analyses also showed ponds of ~0.6-0.7 m were sufficient for nitrate removal, while deeper pond (1.3 m) provided no additional benefit. Long-term total N mass balance revealed 87% of retained TN was attributable to atmospheric N loss, indicating that effective microbial denitrification can develop in cold-region pond-wetland systems, even prior to extensive vegetation establishment or long-term soil accretion. Field-based microbiome community analyses and controlled, anoxic, batch-scale denitrification trails revealed sediments as the dominant habitat of nitrate removal compared to stem-associated biofilms (x2.8 relative abundance of denitrifying taxa and x3 to x9 rate constant normalised to habitat-surface area). The bench-scale results also found denitrification was strongly constrained by temperature at 4 deg C across all habitats, whereas organic carbon availability became the dominant limiting factor at 20 deg C. Overall, this comprehensive study shows that minimally invasive CD and a small footprint pond-wetland system (<1% of farmland) are effective, targeted, implementable engineering BMPs for TSS and N, that provide strong environmental benefits and support broader adoption in cold-climate grain-farming regions.