Estimation in the field of individual perennial ryegrass plant position and dry matter production using a custom made high-throughput image analysis tool

Perennial ryegrass (Lolium perenne L.) is widely employed in grazing systems across the temperate world. Continued persistence of ryegrass pastures depends on our ability to further our understanding of the key traits or attributes responsible for improved plant performance. To this end, plant improvement is highly dependent on precise phenotyping capabilities. However, in-field phenotyping is labour intensive, relies on experienced operators and is often considered cost prohibitive. The aim of this study was to test the accuracy and limitations of an image analysis tool developed to locate and estimate dry matter (DM) production of individual plants growing in the field. Fifty plants were used to develop the image analysis tool and another 1100 to test its accuracy and limitations. DM was measured by weighing oven dried harvested plant material and estimated both with the image analysis tool and by visual scoring. The image analysis tool was faster and estimates of plant area were more highly correlated with measured DM values, potentially capturing 25% more of the variation, than those from visual growth scores. The accuracy of the tool was greatest for DM production values below 15 grams and the tool would benefit from further development to improve estimation accuracy for values of DM production above this threshold.