Reducing phosphorus runoff in snowmelt
Using variable rate fertilizer application in annual grain crops or growing forage mixtures can help reduce phosphorus (P) runoff in snowmelt from drained potholes in the Canadian Prairie Pothole Region.
Drainage management practices such as utilizing surface ditches can help drain water from potholes in the spring to improve crop production efficiencies. However, these practices can also impact water quality downstream.
A research study was implemented to see if fertilizer and crop management practices could be implemented to limit the environmental impacts of pothole drainage in the Canadian Prairie Pothole Region conditions. It was conducted at the Ag in Motion site near Langham, Sask. in a three-year field study starting in the spring of 2021.
The main objective was to see how variable rate fertilizer application, shallow tillage or growing a forage mixture impacted dissolved reactive phosphorus (DRP) concentrations and P load in snowmelt surface runoff from drained watersheds. Crop P uptake and grain yield were also compared for each treatment. In addition, the research looked at how performance varied across different landscape positions on the hummocky landscape.
Eight hydrologically isolated watersheds with ephemeral to seasonal permanent potholes were identified. They covered a total of 16 ha in a single field. In the fall of 2020, surface ditches were constructed to drain the potholes off the field.
In the spring of 2021, snowmelt surface runoff was collected over a four-day period to determine baseline DRP concentrations. Soil samples from 0 to six inch (0-15 cm) were also collected in each watershed to provide baseline physical and chemical properties. These were collected prior to the implementation of the fertilizer and crop management treatments.
The control treatment had a constant fertilizer rate applied across each treatment area and no post-harvest tillage following the annual grain crop.
The second treatment used variable rate nitrogen (N) and P fertilizer rates derived from a commercially available VR prescription plan (SWAT MAPS). The VR plan utilized 10 zones, with seed, N, P, and sulfur (S) fertilizer rates varying across each zone. For example, for Zone 5 in 2023, the seeding rate for canola was 4.8 lbs/ac (5.4 kg/ha), while the fertilizer rates were 100 lbs N/ac (114 kg N/ha), 50 lbs P2O5/ac (56 kg P2O5/ha), and 27 lbs S/ac (30 kg S/ha).
The third treatment was the same as the control, but with a shallow 0 to 2 inches (0-5 cm) post-harvest tillage to lightly incorporate crop residue in the soil surface.
The fourth treatment was an annual forage species (30% Hairy Vetch; 25% Crimson Clover; 25% Tillage Radish; and 20% Turnip) seeded and harvested each year.
The four treatments were randomly assigned to two of the eight watersheds, and applied to the same watershed each year. The three landscape positions sampled in each watershed were upslope, midslope, and depression.
The annual crops grown in the control, VR and tillage treatments were flax in 2021, wheat in 2022, and canola in 2023. All fertilizer was banded at time of seeding.
Lower P losses with VR and forage mixtures
The treatments did not have an impact on grain yield in any of the three years. The VR P treatment rates were 10 to 20 per cent less than those applied to the Control, but were applied to zones with high P fertility which could explain the lack of yield response. The only landscape area with higher grain and straw yields was on the depression compared to the upslope and midslope. The different treatments also did not provide a significant interaction between yield and slope.
Total crop P uptake did not differ in any year for each crop for the Control, VR Fertilizer, and Tillage treatments. The forage mixture P uptake was variable, with higher P uptake than the other three treatments in 2021, but was significantly lower in 2022 and 2023. Landscape only affected crop P uptake in 2021 similar to crop yield where the depression treatment had higher crop P uptake.
Residual soil available P measured post-harvest was not significantly affected by the four treatments or landscape position.
In two of three years, the VR treatment had significantly lower DRP concentration measured in-field surface snowmelt that were 64 to 69 per cent of the control with no significant reduction in crop yield. In two of three years, the Forage Mixture also reduced DRP concentration relative to the control.
Total estimated P load concentrations from in-field snowmelt was significantly lower for the VR fertilizer treatment by 50 per cent compared to the control in one of three years. The Forage Mixture treatment also had significantly lower estimated P load than the control in two of three years.
The differences in P runoff in snowmelt was largely attributed to differences in P fertilizer rates under the VR treatment. The researchers noted that “generally, fertilizer and crop management practices that effectively match fertilizer P application rates to variation in crop P demand across the landscape and that encourage crop removal of residual P can be effective in reducing P transport from drained agricultural watersheds.”
This research was supported by the Saskatchewan Wheat Development Commission, the Saskatchewan Canola Development Commission, the Saskatchewan Soil Conservation Association, the Saskatchewan Stock Growers Association, and Water Security Agency.
B. Weiseth, J.J. Schoenau, and J.A. Elliott. 2026. Fertilizer and crop management practices to reduce phosphorus transport in snowmelt runoff from surface drained agricultural watersheds in the Prairie Pothole Region. Canadian Journal of Soil Science. 106: 1-12. Open Access: https://doi.org/10.1139/cjss-2025-0085
Photo courtesy Blake Weiseth
