Counsel 2: Keli

Project Title

Investigating the Effects of Potato Field Conversion into a Permaculture Orchard on Nitrate Water Pollution in a PEI Watershed.

Overview

Keli's project takes a deeper look at undiversified, monocrop agricultural models present in Prince Edward Island potato farms and how conversion to a polyculture and perennial orchard model could help reduce nitrate runoff pollution. Using QGIS' Soil Water Analysis Tool (QSWAT+) and a variety of water quality, weather, and soil data, this project aims to model watershed output of nitrate into PEI streams. From what I can tell, the relative output of nitrate at various crop-cover percentages will be used as a metric for determining the relative performance of polyculture and monoculture crop systems with respect to fertilizer retention. The "better" case will be the crop system with the lowest overall nitrate output.

Project Objectives

• Model nitrate retention/output in a PEI watershed/stream ecosystem case study with a monoculture crop system

• Model nitrate retention/output in a PEI watershed/stream ecosystem case study with with a polyculture crop system where half of potato field production (i.e., monoculture) is converted to a permaculture orchard (i.e., polyculture).

• Determine the relationship between crop system type (i.e., monoculture vs. polyculture) and riparian nitrate output via agricultural runoff.

Data

Keli is largely relying on Environment and Climate Change Canada (ECCC) and Environment Canada publicly available open data portals for water quality, weather, and soil data. In order for Keli to perform the methodology outlined in her project, including performing a SWAT+ analysis in QGIS, she will need to acquire a DEM of sufficiently fine resolution to be able to delineate watershed and stream systems and identify inflow and outflow points. She is relying on her meteorological and soil data being both high quality and usable for developing a comparison of nitrate outputs under the monoculture and polyculture crop system models. Another consideration with respect to data is the temporal extent under which she will be running her model. This inevitably requires that her data span a sufficiently long time period under both model cases.

Methodology

By incorporating hydrological response units related to slope/topography, land use proportions, soil data, precipitation, temperature, relative humidity, solar radiation, and windspeed, Keli aims to determine the relative quantity of nitrate leaving the system via riparian systems. The first model will be run with land use proportions reflecting the monoculture crop system seen in current practices with PEI potato fields; the second model will be structured similarly, though land use proportions will be altered to reflect the polyculture crop system that would be in place if half of potato field production were converted to a permaculture orchard. 

In order to make this adjustment for the second model, she will need to assign a land use percentage to each pixel of the land cover data. By doing so, she will be able to select half of the existing potato land cover and replace it with the polyculture class. This will likely involve using something similar to the Raster Calculator tool, as well as the Raster to Polygon tool, available within ArcMap. I am not sure if this holds true for QGIS, but I know that the Raster Calculator tool available within ArcMap is limited in that it can only be used within Model Buidler and cannot be adapted to ArcPy, the software's python shell. If this is also the case within QGIS and Keli intends to make use of the software's programming interface to perform more complex analyses, she will need to find a workaround.

After exploring her project proposal, I am left with a lingering question regarding how Keli intends to compare nitrate outputs from the two models. Are conclusions regarding the viability of the monoculture/polyculture systems going to be based solely on empirical comparisons between nitrate outputs of the two models? Does she intend to perform some sort of statistical analysis to bolster the significance of her results? Though these considerations are likely not necessary for the scale of the project at hand, it would certainly make Keli's final results more robust if they were run through some sort of post-processing analysis.

Possible Problems and Considerations

• Are conclusions regarding the viability of the monoculture/polyculture systems going to be based solely on empirical comparisons between nitrate outputs of the two models? What kind of statistical analyses, if any, will be employed?

• Over what temporal extent do you intend to run your models? What impact, if any, does time have on the final nitrate output of each model and your results overall?

• How does the SWAT+ tool actually work? It would be beneficial to your readers to understand exactly how each of the water quality, meteorological, and soil parameters you mention in your proposal will be fed into the tool and used to develop your final nitrate output data for each model.