Harmful algal blooms are caused by excess phosphorus in many parts of the world. In most years, a few peak flow events, such as high rain or snowmelt, have caused the majority of phosphorus runoff from agricultural lands to nearby watersheds. In cooler climates with significant winter periods – such as those in Canada and the northern United States – runoff occurs primarily during the non-growing season (October – April). Yet, little is known about how phosphorus behaves during peak flow events during this time. At present, most of our understanding of runoff drivers and pathways has been gained during warm periods and the growing season (May – September). In particular, there is a lack of detailed year-round field-scale studies that report on both of the following and how they influence one another: 1) the buried networks of drainage pipes on agricultural lands, known as tile drainage systems, and 2) the occurrence of overland flow during the non-growing season. These gaps in our scientific understanding hinder our ability to advise farmers and water managers on how to best manage phosphorus pollution from agricultural lands.
Methodology
This paper is part of the Master of Science thesis work of University of Waterloo student Chris Van Esbroeck. He became aware that frequent logistical challenges or equipment failures have left scientists struggling to acquire data on hydrology and water quality throughout the non-growing season, primarily due to freeze-thaw cycles and the formation of ice on and around equipment. Together with our team, he was able to overcome these challenges by using a combination of specialized equipment and insulated enclosures, and by taking advantage of the warmth of the ground to insulate equipment where possible.
In consultation with partners from the Ontario Ministry of Agriculture, Food and Rural Affairs, we designed this study to measure hydrologic and biogeochemical phosphorus losses in surface runoff and tile drainage from two working farms in Southern Ontario, Canada: one in Ilderton and the other in Londesborough. Both sites had comparable tile- drainage infrastructure with standard, plastic pipes and shared similar environmental and management characteristics. Surface runoff and tile drainage from both fields exited predominantly in one location, allowing the measurement of edge of field runoff.
At each site, a section of the main tile drain was removed and replaced with a custom-built piece that was used to measure flow and collect water quality samples. Surface runoff was captured in a common outlet using berms and a culvert at one site and a surface inlet at the other site. Tile and surface runoff monitoring stations were equipped with a set of automated water samplers triggered by water depth responses. Both particulate phosphorus (attached to soil or sediment) and dissolved phosphorus (what remains in water after that water has been filtered to remove particulate matter) concentrations were measured.
Our study was conducted year-round, but emphasized peak flow events that occurred throughout the non-growing season. We relate inter-event variability in the quantity and speciation (dissolved or particulate) of phosphorus to event climatic drivers (e.g. rainfall, rain-on-snow, and snowmelt) and pre-event soil conditions (e.g. presence or absence of snow cover in presence of frozen ground).
Outcomes
Most runoff and phosphorus losses occurred during 5 to 6 peak events during the non-growing season, particularly during the major snowmelt period (Fig. 1). During these peak events, we saw both surface runoff (or significant surface ponding) and tile drainage, yet the surface runoff had greater and more variable concentrations of phosphorus (Fig. 1). We also found most phosphorus losses (70–90%) were in particulate form, and were greatest when rain fell on thawed, bare soils. In contrast, dissolved phosphorus losses appeared to decline throughout the winter season.
Figure 1: (a) Daily total precipitation (left axis) and snow depth on ground (right axis), (b) mean daily air and soil (10 cm depth) temperature (right axis) and water table position below the surface (left axis), and (c) total daily discharge from tile drains and as surface runoff from the Ilderton (ILD) site. Peak flow events are identified in (c) with arrows.
Conclusions
With most phosphorus losses coming from a few peak events where overland flow is generated, we are able to see the problem is not about chronic, longer-term event cycles. It has been previously documented, by my research team and others, that overland flow is the dominant source of phosphorus losses, but this study tells us specifically how tile drains factor into the total phosphorus losses at the edge of the field.
When designing best practices for managing runoff from agricultural lands, farmers and water managers need to be informed about how much tile drainage contributes to phosphorus runoff because it impacts the resources they put into blocking drains and tilling their soils. Knowing that most phosphorus runoff, during the non-growing season, comes from surface runoff and not tile drainage will help farmers and water managers prioritize the best management practices that they use, and will help them make decisions on how to manage fields in autumn, prior to the winter season.
Anticipating elevated concentrations of particulate phosphorus during rain on bare soil events will also encourage them to leave residue in place to prevent erosion. The use of residue to protect soils in winter may become more important under a warmer climate if there is more rain and less soil frost. More research, however, is needed to determine if the potential loss of dissolved reactive phosphorus from residue indeed offsets the ability of the residue to mitigate particulate phosphorus losses.
Following the publication of this study we have already begun to examine how much and where to apply phosphorus to crops and how to use cover crops to mitigate phosphorus losses.
Insight into the amount and timing of edge of field losses can assist managers with setting targets, developing management strategies, and quantifying loads in the tributaries of the Great Lakes. In addition, an improved understanding of runoff and biogeochemical pathways, and forms of phosphorus will improve existing models, which currently do not effectively simulate subsurface phosphorus loss or winter processes in general.
Overall, the characterization of runoff and phosphorus load responses under a range of non-growing season conditions and event types will improve our understanding of this critical period and will assist us in predicting how water quality issues may change with climate change. Further, such information can provide insight into the identification of best management practices that may be most effective through the non-growing season.
Van Esbroeck, C.J., Macrae, M.L., Brunke, R.R., & McKague K. (2017). Surface and subsurface phosphorus export from agricultural fields during peak flow events over the nongrowing season in regions with cool, temperate climates. Journal of Soil and Water Conservation, 72(1), 65-76.
Contact: Merrin Macrae, Department of Geography and Environmental Management
For more information about WaterResearch, contact Amy Geddes.
