Each week of the growing season begins the same for Shea Thomas and David Athey of Baker Crop Consulting LLC. They go to flagged locations in each of the major soil areas of cotton fields to map plants.
Information they gather at the sites are fed into an industrial grade form of the crop simulation model expert system, Gossym Consultant. Among other things, it has the capability to convert weekly boll counts into plant maps.
What is so unusual about this is the three scouts plant-map about 17,000 acres a week, spending around five minutes at each sampling site, a fraction of the time usually spent for a complete plant map.
The speedy plant mapping technique is just one of several new developments making precision agriculture a reality on a handful of Mississippi farms.
Other new technologies to agriculture include aerial and satellite multi-spectral imagery, which are replacing yield maps and grid sampling as cornerstones of precision farming.
The showplace for much of the research on precision agriculture is Perthshire Farms, run by the Hood brothers in Gunnison, Miss. About 12,000 acres of cotton there are under a precision agriculture program.
Crop consultant Don Baker stresses that precision agriculture on Perthshire Farms is far ahead of where it was just a couple of years ago.
New technology has something to do with this. But how the technology is delivered and used by the farmer is changing, too, according to Baker. "Growers have been getting into precision farming with the expectation that their yield monitors will tell them where to fertilize and how much to fertilizer. But it hasn’t worked as well as they had hoped."
One reason — cotton is a management-intensive crop that demands weekly adjustments to a cotton production program. For example, "A farmer doesn’t necessarily want to know how much nitrogen it’s going to take to raise cotton," Baker said. "He wants to apply about half of it at planting and then he wants to hedge his bets to see what happens the rest of the season."
New scouting techniques, petiole analysis, computer modeling, variable-rate application technology, and aerial and satellite imagery are making the latter possible.
It begins with weekly scouting. Instead of doing a complete plant map at each site, Baker’s scouts take data on plant height, node numbers, nodes to the first fruiting branch and boll counts.
The data are fed into Gossym Consultant, which generates the plant maps and updates the progress of the crop.
The update is about as sophisticated as you can get. Gossym has stored about 160 profiles in its computer banks. The profiles, also called management units, have common factors, such as soil type, planting date, irrigation, irrigation method, etc.
A crop developmental trajectory generated for each of the profiles predicts what day the crop in that profile will reach a certain stage of development or begin to stress — bloom, cutout, nitrogen and moisture deficiency, etc.
But as cotton producers know, each week is different. so various hypothetical future weather scenarios are considered.
Every Monday, Gossym Consultant generates a projection report, which gives the cotton producer information for two scenarios, a hot season and a cool season. The report includes eventual plant height, when the next irrigation will be needed if there is no rain, the amount of additional nitrogen needed, defoliation date and a projected yield.
The projections are "all based on this raw data that is coming in from this weekly report on plant height, nodes, your plant maps," Baker said.
Every two weeks, leaf samples are taken "to tell us whether we are right or not with our projections on fertility," Baker said. "If we are wrong it’s because we did not estimate the carryover as well as we should have."
The program will tell the Hoods when the plant will run out of fertilizer and calculate the yield increase from making an application in a specific management unit. The plant height reading provides insight into when and where the Hoods should make a Pix application. "We’ll also solve for the amount of Pix that’s needed," Baker said.
Interestingly, the computer may decide that a Pix application, although needed to reduce plant height, would not be a cost-effective application. "The computer tells us not only when and where to apply, but sometimes whether to apply," Baker said. "Because there are many developmental trajectories that may lead to any particular yield, our projections are far more powerful than simple yield monitor data.
"The report also tells the producer when he should have those pickers ready to go and the order of defoliation."
Of course, Perthshire Farms has the variable-rate equipment necessary to treat each management unit according to it own needs, maximizing efficiency and yield. Sometimes, there may be as many as four or five different management units in a field.
Multi-spectral imagery also provides excellent in-season clues on the development of the cotton plant, helps verify what the computer model is projecting, and guides scouts to areas where problem may be arising, noted Baker.
The potential for imagery is unlimited at this point, Baker noted. For example, "Once rules can be established on the feeding habits and behavior of specific pests, the imagery can tell the entomologist where to look for those problems in a field.
"The work of USDA/ARS scientist Jeff Willers shows that you don’t look for some pests in the less-vigorously growing part of the field. It’s much less attractive to that type of insect."
Baker believes that precision agriculture on Perthshire Farms is just the revolutionary system of management that cotton producers need to reach new yield plateaus.
In fact, it almost happened last year on the Perthshire Farms. Site-specific farming techniques were used on the farm to produce what Baker described as the highest potential yield ever on the operation.
But rains began to fall just as bolls were cracking, which resulted in severe boll rot. Yield was significantly reduced although Perthshire still ended up with an average crop.
Precision farming "is so multi-faceted," Baker said. "There are many opportunities. We really haven’t worked with it enough to know what we can do with it. We’ve talked about saving chemicals and saving seed. But we haven’t talked about finding those areas that can yield four bales and getting those four bales rather than the two we would get if we averaged our inputs over the whole field."
Multi-spectral imagery can help a producer identify that high-yield potential.
"There is the potential to really blow the lid off and go for high yields in those areas capable of it," Baker said.
Precision farming could help the producer get there "by doing everything right," Baker adds. "There are dozens of paths to an 800-pound yield, but only a few paths to a four-bale yield."