Research HighlightsLeaf-specific application of herbicide
By Kentucky Soybean Board
The Kentucky Soybean Promotion Board funds a wide variety of research projects designed to benefit soybean farmers in our state and beyond.
One such project, currently in its second year, is being conducted by Dr. Robert Pilgrim of Concurrent Solutions, LLC. During the project’s first year, Pilgrim established a baseline by demonstrating the efficacy leaf-specific application of systemic herbicide (e.g. glyphosate) on weeds of concern to soybean growers.
This second year entails greenhouse demonstrations of targeting accuracy of leaf-specific herbicide application with minimal transfer to crop plants. Year three will include field testing of proof-of-principle leaf-specific herbicide application in non-herbicide resistant soybean crops.
Soybean producers will benefit in a number of ways if Pilgrim’s application method proves to be viable. They’ll see significant reductions in the amount of herbicide necessary to control weeds in soybean crops, post-emergence. They’ll gain the capability to apply systemic herbicide to weeds, post-emergence, on non-GMO soybeans. A decrease in stress to crop plants is a natural benefit of weed-only herbicide application, and a yield bump generally accompanies a decrease in stress.
High-resolution imaging of the crop plants offers side benefits of detection of other threats to the crop plants such as insect and disease, along with providing data which will help in the accuracy of early yield estimates. Finally, this project may provide future opportunity to apply more effective herbicides to weeds that have developed a resistance to glyphosate or other herbicides for which GMO crops permit post-emergent spraying.
Pilgrim and his colleagues have been working on the development of what he calls the “next generation of precision agriculture” since 2003, specifically imaging and control processes at a resolution of millimeters. Pilgrim says that his system will demonstrate a capability to distinguish between soybean crop plants and weeds using real-time machine vision processing and determine the efficacy of application of glyphosate to weeds of concern to soybean growers.
Pilgrim and his partners have devised an overhead rail system, which emulates on a small scale the towed or autonomous herbicide application system, testing weed/crop classification, machine-vision, weed targeting and selective herbicide application. The rail system provides the ability to control experimental variables that would be difficult or impossible in the field and to repeat trials to gather data on performance and to modify, refine and test critical hardware components.
The machine vision systems developed in the first year will be tested in the greenhouse to support modification and demonstration of existing machine vision algorithms for soybeans and weeds of interest. As was done in a previous study in the United Kingdom, researchers will collect and analyze a wide variety of weed images in the field with natural plant and clutter backgrounds.
Early stages in the machine vision processing perform a segmentation of the images into separate objects. These segmented objects will be evaluated visually to determine “ground truth” for plant type and each “truthed” object will be tagged to support functional performance testing of the machine vision algorithms as they are developed.
As herbicide-resistant weeds continue to develop and compete with crop plants for nutrients, water and sunlight, increased modes of action for removing these weeds are needed. Many weed scientists advocate farmers going back to the primitive strategy of manually walking their fields to remove and therefore control these weeds.
For a ten-acre hobby farm, walking the fields is a reasonable component of an intensive weed management. For an increasing number of farmers – those who farm thousands of acres –that recommendation is not only difficult to comprehend but impossible to implement. Pilgrim’s vision of an apparatus similar to an irrigation pivot (only armed with herbicide and using precision ag technology to put that herbicide only where it is needed) could be a welcome addition to the toolbox that farmers are building daily in the fight against herbicide-resistant weeds.