Innovations in Agriculture

 

On December 3rd, several Ohio speakers discussed weather, Artificial Intelligence (AI) in agriculture, and precision soil sampling at the Ohio No-Till conference, Der Dutchman, Plain City Ohio. Here are their comments.

First, Dr Vinayak Shedekar, OSU Ohio Drainage Management specialist spoke about weather, mainly drought and excess water. Nationally based on insurance claims; drought/dry weather cost farmers about $90 billion/year (43% of total) followed by excess water $52 billion (25%) with other insurance claims around $22 billion/yr (32%). Water is one of the most limiting factors to obtaining high crop yields. In Ohio, typically 70% of water related crop losses are due to excess water with 30% due to drought. Recently though, drought has been a bigger concern. Drought may cut corn yields 40 bushel/acre or more while typically too much water historically reduces corn yields 20 bushel or more.

Corn needs about 28 inches of rain during the growing season (Late April to Late September). Typically, during the winter and early spring, farmers get too much water followed by too little in late spring, summer, and fall. July and August typically have a water deficit of 3.6-4.0 inches. The best way to make up that deficit is with good soil structure and high soil organic matter (SOM) levels. A farmer shared that he planted through a long-term undisturbed soil that was part of a waterway and at harvest, due to the good soil structure and high SOM levels, his yield monitor jumped to 359 corn bushels/acre.

Dr. Scot Schearer, OSU Department Chair for FABE talked about how AI. Artificial Intelligence (AI) is changing how farmers apply herbicides, especially post emergence, reducing herbicide usage 60-70% by spot spraying. AI is identifying nutrient deficiencies, crop diseases through visual images and helping farmers make better decisions to lower fuel costs. Two companies, Bosch & BASF (a herbicide company) have teamed up to use less herbicides which is safer for the environment. BASF is pushing services to make up the difference and also raising prices. John Deere is investing heavily in corn-soybean-wheat- and cotton AI research. Today with image collection, 6 to 7 terra bytes of information can be collected daily. A typical phone collects 1 terra byte of data (one trillion data points).

Dr Schearer says the era of large machines is starting to peak and go the other way. Small autonomous equipment have much lower labor costs and can operate 24 hours per day. Autonomous (self-driving) sprayers are 300# of efficiency and simplicity according to Dr. Schearer. About 30% of all fungicides on wheat this year was applied with drones. Dr. Schearer’s son lost his AI job to another farm company. He was reluctant to working on a farm in California until he discovered that this farm is a $7 billion dollar business producing nearly all the lettuce in California. Jobs will change but people can adapt to doing different work; especially work that computers, robots and drones can do better that is repetitive and back breaking.

Dr. Schearer says all farmers can use AI to change agriculture. On CHAT ID, farmers can take a picture of a yellow soybean plant and CHAT ID will describe the symptoms, tell you the problem, and offer a solution. Example: Your soybeans have yellowing between the veins and it occurs on the younger leaves. This indicates manganese, iron, and sulfur deficiencies and early signs of SDS (Sudden Death Symptoms). The program then recommends varies sprays and solutions, the cost, along with timing for best results.

The third speaker was Dave Scheiderer with Integrated Ag Services (IAS) who talked about using precision soil sampling to reduce input costs and increase crop yields. IAS has a machine that can take 1 sample about every 15 seconds or every 100 to 150 feet in the field. They are grid soil sampling on 1/8 up to 2.5-acre grids. They find 1-acre grids or smaller give the best results, depending upon the nutrient. They take samples at precise depths because nutrient variability decreases greatly with depth. This reduces variability in their results. Mostly they take 4-inch samples and they often sample 2X a year. The biggest returns come from adjusting pH (lime applications) and potassium (K) soil nutrient levels.

Sampling one acre costs $11/A per year and takes 40# of K or 500# of lime savings to pay for itself. K varies the most ranging from 0-12% variability. K soil levels are highly influenced by soil texture (sand/silt/clay levels), soil compaction, soil health & biology, weather (how dry or wet), existing soil nutrient levels, and other factors. By sampling frequently and over small grids, they increase the accuracy of their data and recommendations.