In 1998, the first year of the IHARF/SSCA Centre of Excellence for Precision Farming Research, a canola - spring wheat - field pea - spring wheat rotation was established by separating the 308 acres into eight fields so that each crop was replicated twice. Infra red and remote sensing images were collected as well as yield maps. 308 permanent geo-referenced grid points were established and permanent pins inserted in the soil at each point in the spring prior to seeding.
In the fall of 1998, soil samples were collected at each point for a detailed soil nutrient map and a full soil salinity geo-referenced map was made of the entire site. In the spring of 1999, a second salinity map was done and further soil samples were taken at selected sites in order to determine soil moisture levels as well as measurements of surface roughness and crop residues. Measurements were taken to try and correlate their values to data collected from the Radar Sat. During the growing season, plant and weed counts, disease assessments and collection of grain samples for nitrogen determination were done around each pin. Additional infra red and satellite images were also collected and all fields were yield mapped at harvest.
Funds from Agri-Food Innovation Fund (AFIF) and Canadian Adaptation and Rural Development Saskatchewan (CARDS) allowed us to expand research at the site in 1999. The AFIF funds allowed us to hire a full time technician and a summer student, acquire more remote sensing images, survey of the site with a salinity meter (EM38) and establish some measurements for the monitoring of deep nitrates. The CARDS program gave us the opportunity to conduct a comprehensive weed and disease survey and hire Paul Bullock, Noetix Research Inc. to analyze the information that was gathered. This analysis will help to assign management units to each field and from there to develop an appropriate variable rate application fertilizer program. Slide pictures were taken of every quadrat where weed counts were done. This will allow us to do image analysis on each quatrat and to determine the accuracy of image analysis for doing weed counts. The image analysis is being conducted by Dr. Trever Crow of the University of Saskatchewan and his graduate student, Harry Ingleby. Another component of the proposal is the feasibility of using robotics to do weed scouting in the field. This is coordinated by Ron Palmer of the University of Regina with his graduate student, David Wilke.
In 2000, a detailed prescription map for nitrogen requirements will be done on the site in order to establish a variable rate map. The extensive collection of data will help us identify cost-effective ways of assigning management units to a field without having to resort to extensive soil sampling. These units will allow us to more fully understand the variability in a field and determine how to work with this variability to achieve maximum profitability.