A lot of my experience in the precision farming area has been gained through work with Vance and Fran Simpson. They farmed 34-quarter sections of land near Raymore, Saskatchewan this past year. Initially, conference organizers wanted Vance to speak to you here today, however they are currently away in Yuma, Arizona. That in itself probably attests to the success of their farming operation using precision farming methods. Hence, this is why you're stuck with me here today.
I have been fortunate to be able to find a good farmer cooperator to work with, and we "learned the ropes" together. It's not that easy to come across willing cooperators that have the full line of precision farming equipment.
The Simpsons started into precision farming in 1977 with the acquisition of two new Case IH 2188 combines equipped with the full AFS yield monitoring system. The local Ag. Rep., Lewis Reeve was initially consulted about this equipment, and that is how I got involved. Hence yield data has already been gathered over five growing seasons. With proper calibration, checking actual weights of grain harvested, the Simpson's are able to achieve very accurate data with their yield monitors.
After getting an accurate picture of varying crop yields in their fields, the Simpson's were anxious to use this technology to improve their bottom-line. Early in 1998, the Simpson's purchased a Flexi-Coil 50 series air cart with the variable-rate option. This allows an operator to automatically change application rates "on the go" in a field, according to a prescription file on a card in the controller in the tractor. The GPS receiver from the combine is moved to the tractor for seeding to give the unit its exact position in the world.
Now came the challenge of writing a prescription and deciding on what to tell the variable-rate applicator to do in field. Because fertilizer (especially nitrogen) is a costly and influential input in continuous crop production, initially the focus has been on variable rate fertilizer technology -- varying the rate of nitrogen fertilizer across a field.
Originally, after consultation with the soil science department at the University of Saskatchewan, the criteria for variable rate fertilizer technology was topography or landscape within a field. Knolls have differing production capabilities from those of depressions. We found that it was not very simple to get a map of a field showing the different landscape positions -- knolls and depressions. The University of Saskatchewan soil science department provided a quick and economical way, using image analysis. Computer software scrutinized a black and white air photograph. Lighter colored areas corresponding to knolls were separated out from darker colored areas corresponding to depressions. Initially, fields were divided into three zones -- upper, mid slopes and lower landscape positions. This resulted in a large number of zones for each field, over 200 in some cases.
For the first year of variable-rate prescriptions, a rather novel method of benchmark soil sampling was used. After studying the soil types on a field, three benchmark sites were picked representing a typical knoll, mid slope and lower slope. A separate soil test was done for each of these three points and the resulting recommendations were extrapolated to other knolls, mid slopes and lower slope positions for that field. I wrote a simple spreadsheet program to calculate fertilizer blends and rates for the various management zones. Prescriptions were written for 6-quarter sections using this method in 1998. Fields were divided up into five or six plots to provide a good comparison of variable-rate to conventional constant rate fertilization. Plant tissue tests were done to verify adequate nutrient status.
The first year's results were encouraging, although results were variable. Yield differences between knolls and depressions were fairly dramatic, especially in canola, with yields in depressions usually much higher.
In one field of barley, a lot more nitrogen was applied on the knolls and yields were higher than moderately fertilized depressions, however this was an exception. We found that generally it was not possible to push yields higher on knolls than in depressions. The depressions have a lot more inherent organic matter, soil fertility and moisture, thus having a greater yield potential.
In one field, saline areas were defined and mapped out. The prescription was written to turn off fertilizer application in these areas. This system has vast potential for reducing crop inputs and avoiding waste.
After reviewing the results of the first year of trials, the Simpson's changed their methodology of variable-rate fertilization, again based on the recommendations of soil science researchers at the University of Saskatchewan. For the 1999 growing season, fields were divided up into different management zones -- only two of them -- uppers and lowers. This made for a much simpler system of about 20 zones per quarter section. Soil testing was carried out using conventional methodology, i.e., sampling the field in a minimum of 15 locations at mid slope positions. The recommended nitrogen rate was then applied to the upper management zones, and depending on the crop, 1.5 to two times the recommended nitrogen rate was applied to the lower slope management zones. In 1999, a year of good soil moisture, crop yields were pushed significantly, especially in canola on lower slopes. This same type of variable rate fertilizer regime was used on 11-quarter sections over the last three growing seasons.
Three out of the last four years had good growing season moisture, but 2001 brought a year of drought stress. The barley yield in field #3 was 74.5 bushels per acre in lower slope positions and only 25.9 bushels per acre in the upper slopes. That's almost triple the yield where more moisture produced a response to the higher rate of nitrogen applied.
A detailed explanation of the activities in the early years of this project was contained in the April 2000 edition of Farming Magazine.
For 2001, the Simpson's switched their high disturbance direct seeding operation to zero till seeding with the purchase of a 57-foot Flexi-Coil 5000 air drill. Also in 2001, the two Case 2188 combines were traded for 2388's, with the newer style yield monitors. So they must be doing something right!
Looking at other field trials involving variable-rate nitrogen application based on field topography or landscape position, this method seems to be only moderately successful. It has worked well for the Simpson's, especially with canola.
Precision farming or site-specific field management is a new and emerging technology in crop production. Many large operators, young farmers and innovative producers are starting to use these tools. As combines are being replaced, many new machines are being purchased with yield monitors. Variable rate application equipment is also making its way out into the countryside.
The technology is fully developed and works real well. However, the problem is linking and applying this technology to the agronomy of crop production. Some limited research on this topic has been carried out in this province, however the practical extension of this research in terms of application at the farm level is seriously lacking.
The whole idea is to try and improve the "bottom line" for grain producers. The key to higher profitability is to delineate appropriate management zones in a field. Nitrogen is an influential input in continuous crop production. There should be a huge potential in site-specific management, i.e., varying the rate of this nutrient across a field using variable rate technology (VRT).
This approach has been tried on a limited scale on a few Saskatchewan fields. It would appear that there have been some successes and that the concept has merit. But it needs to be more thoroughly tested at the farm level.
Following the lead of research scientists, management zones were first established using topographical criteria - upper and lower landscape positions. However, there are other ways to delineate management zones. These methods rely on "Mother Nature," and look at past performance or variations in crop growth throughout a field. The first method is to look at satellite images of crop growth in that field over several years. Areas of good and poor crop growth can be determined, and variable rate technology can be implemented based on those management zones.
The second method is at this time a very critical answer to inquiries from producers who have yield monitors and are asking what they could do with their years of yield data and maps. This method involves detailed analysis of three or more years of yield maps on a specific field and is called multi-temporal yield analysis. Although different crops have probably been grown over the three years of yield data, this analysis will point out areas of the field that have consistently produced lower and higher yields, and the areas that have experienced a great deal of change in their performance over the years. This analysis can provide meaningful information about the performance of a producer's field over time, regardless of the type or variety of crop that was grown there. This tool allows one to quantitatively express the differences in a field's yield across time and across the field as well.
There is currently other research under way to improve the predictability and performance of variable rate fertilizer applications. All of these concepts need be tried at the field level. I would personally like to do some "ground-truthing" of management zones by hooking a hand held computer with appropriate software to a GPS unit. The industry is on the verge of releasing a protein monitor. One of these instruments would go a long way towards unlocking the mystery of varying wheat yields.
So what have we learned in terms of the practical use of precision farming methodology at the field level?
(1) Crop yields are showing a lot of variability within a field.
(2) It is not always easy to explain the cause of the yield variability.
(3) Yield monitors are great agronomic tools. They give an instant indication of yield fluctuations from things like weed pressure, insects, soil variability, moisture, salinity and topography.
(4) Precision farming has unlimited potential for doing one's own agronomic trials on the farm. Accurate results can be obtained quickly.
(5) The equipment is great for marking things like stones and weed patches in fields.
(6) Data is easily gathered for analysis of variable rate fertilization. The concept of variable rate application of nitrogen by field landscape position has been moderately successful.
(7) We will be looking at two other methods of delineating management zones based on crop growth of past years -- looking at the work of "Mother Nature." It's not always easy to predict what will happen.
(8) This system has the potential for reducing crop inputs and eliminating waste.
(9) All this technology works well, however the agronomy has some catching up to do.
(10) The yield files are a great data bank for crop insurance and field production records.
The goal of Vance and Fran Simpson was to net an extra $10 per acre using precision farming technology, by increasing yields or reducing fertilizer inputs. To date, they haven't increased their net revenue as much as they would have liked, but have gained a lot of valuable experience and knowledge. I think they are fairly close to their goal. They are continuing to work toward a more profitable net return.
In closing, I'll leave you with one thought.
Most producers are still applying one constant rate of fertilizer to square blocks of land (namely quarter sections) that were surveyed in this province starting in 1877? Maybe there is a better way.