"Down the road, precision farming software is going to provide an economic anlysis that shows me where to invest money in my farm for the best return. Like a retail store that maps out its profits on a square-foot basis so it can put high-profit products in high-traffic centers, GPS field information will help me create management zones and make better buying, production and selling decisions. We're not there yet, but it's coming." - Curt Watson, precision farmer, Renville, MN
Some have likened precision farming information to a new crop. Farmers are busy harvesting the new "information crop" with yield monitors and satellite and aerial imagery, but few know how to use it to make a profit.
The process of analyzing and interpreting precision data is far from easy because crop production is a dynamic, complex biological system. An area of the field that is high yielding one year often is low yielding the next year. Frequently, fertility levels, soil types and yields show no correlation to each other. Weather, insect and disease pressures vary each year and cannot be predicted. Althoughthe goal of information harvesting is to gain more control, it often simply creates more confusion.
Crop consultants and farmers are struggling to find a way to create useful management zones within a field. Foot-by-foot variability is fine for variable rate applications, but a broader perspective is useful and more economical for other management decisions. "A management zone might range from two acres to 15 acres depending on the practice or input you're managing," says Bernie Paulson, McPherson Crop Management. "Five zones per field is a manageable number."
There is no one right system for determining management zones. And every system will likely evolve because this new field of study challenges all previous assumptions. Here's how some crop consultants and their clients are determining management zones.
GeoFARM's approach. GeoFARM in Lake Crystal, MN, is a precision farming data servicing center owned and operated by a consortium of independent crop consultants. Paulson is one of the owners, as is Maggie Jones of Blue Earth Agronomics in Lake Crystal, MN.
GeoFARM geo-references yield maps, soil surveys, aerial maps, history, management information and in-season observations. "Through a cross-classification process, we identify areas with stable production and those with unstable production over multiple years," Jones says.
GeoFARM uses SST Toolbox software to help identify the zones because it can compare several layers of data at one time. "We define high, medium and low yields statistically using standard deviation from the mean yield. It's an objective way of doing this and also factors in yield variability due to annual weather differences," Jones adds.
This method results in three stable production zones: consistently high yielding, consistently medium yielding and consistently low yielding. These zones are stable across both corn and soybean crops over multiple years. It also identifies two unstable zones. They represent areas that have low bean yields and high corn yields or the inverse: low corn yields and high bean yields. "We use five colors to define these five zones on a map. You quickly see the areas and can look for ways to manage them differently," explains Jones. "One client had an area that had extremely high corn yields, but the soybean yields in that area were terrible the following year. The weather was perfect both years. We knew there must be some reason why the beans were doing poorly. We sampled for soybean cyst nematodes and found they were high in that zone."
Larson Systems' approach. "Our primary business is providing data processing services for the farm, allowing the farmer to concentrate on management issues in his operation and not get bogged down with analysis," says Don Larson, president of Larson Systems, Ames, IA, which processes precision farming data from 250,000 acres throughout the country. "It's easy for farmers who don't understand statistical analysis to be sold a bill of goods. They buy software or services that give them a pretty picture instead of an analytical tool to tell them what is affecting the crop in that field. Most people have no idea what is statistically valid and what is not valid.
"Too many people do variable rate application of fertilizer without first building a three- to five-year database of historic performance of the field. Without this site-specific information, it's impossible to determine what benefit they derived from the variable rate application," he says.
According to Larson, fertility mapping and variable rate application of fertilizer are two of the least influential aspects of crop production and two of the least cost-effective uses of precision farming information. He says, "We haven't seen a benefit to variable application of fertilizer or any basis for intensive soil sampling. Overall, the lowest-yielding parts of a field have the highest fertility, and the highest-producing areas have the lowest fertility. There's not a payback to adding more fertilizer. The crop is doing well because the natural environment of the soil allows it to thrive. It's much more beneficial to use precision tools to study the naturally occurring characteristics of soil, to make water and drainage management decisions and to compare genetic performance of varieties based on tillage practices and types of soil."
To define management areas, Larson Systems starts by looking at the soil characteristics in the field. "An accurate Soil Conservation Service soil map and yield data are the baseline. As you collect yield data over time, you start to see good production areas and poor production areas. Quite often they correlate to different soil characteristics," says Larson. "Typically our management zones look like the soil maps. Soil structure, organic matter, pH, water-holding capacity - these are the primary factors affecting yield. The county soil book is farmers' most valuable resource and it's collecting dust on the shelf.
"We've been helping farmers define management zones since 1991 and we consistently generate a $25 to $50 increase in net profit for them," says Larson. "The biggest payback has come from tiling and variable rate planting by soil characteristics."
Midwest Ag Services' approach. Midwest Ag Services has three locations in North Dakota - Jamestown, Grand Rapids and Steele - and one in Huffton, SD. Glenn Hanson, director of the company, has spent four years developing a directed soil sampling system for dryland, small grain production. "I knew our small grain growers would not spend $15 to $25 per acre per year on soil sampling in fields where they spend only $15 to $25 per acre in fertilizer inputs for wheat, barley or durum," says Hanson. "I had to reduce sampling costs and I felt creating soil-sampling management zones for nitrogen would do that. If it was profitable for small grains, it would work for other crops as well."
Hanson conducted his own research to verify work done by Dr. Dave Franzen, extension soil specialist at North Dakota State University, that showed a strong relationship between landscape topography and placement of mobile nutrients (nitrogren, sulfur and chloride) in the field. Franzen's work compared 11/10-acre grid sampling to sampling by management zones created from topography and soil conductivity maps. Sampling by management zone was more reliable and effective than grid sampling. "We were able to take one-sixth the number of soil samples, six samples versus 36 samples, in a 40-acre field and still maintain a high level of confidence in the information with relatively low cost," explains Franzen.Here's how Midwest Ag Services creates the topographical management zones covering 25,000 acres. It uses a Trimble 4700RTK GPS system to create elevation maps accurate within 2 cm. At the same time, it pulls a Veris 3100 sensor to map soil conductivity in the top foot and top 3 ft. of soil. The firm developed software that integrates data from both pieces of equipment and creates field boundaries. Up to 12 zones can be identified. Generally five to nine management zones are identified within a field.
Even in relatively flat landscapes like the Red River Valley, small differences in elevation significantly affect the movement and pattern of nutrients in a field. Denitrification occurs in depression areas where water pools. Soil conductivity measurements often are related to soil texture. This information brings more detail to soil survey maps and can help predict areas of nutrient leaching. "Soil survey maps were never designed to direct variable rate fertilizer applications," says Franzen. "Soil inclusions that can account for 30 to 40% of the field do not show up in a soil survey. The conductivity data help us better define soil characteristics on a site-specific basis."
"We measure as many layers as possible and then look for the parts of the individual layers that correlate to nutrient retention and crop production and only use those layers to create the management zone," says Hanson. "If you try to weight management zone creation on one factor, you'll fail."