Development of Remote Sensing Tools for Alfalfa Management.

Accurate, in-field assessment of alfalfa yield, quality, and persistence are critical to optimize production and resource-use efficiency. Modern remote sensing technologies (i.e. measurement of canopy reflectance) have potential to facilitate accurate predictions and account for spatial variability, although alfalfa specific management tools have not been developed. A study was initiated in 2014 at the University of Minnesota Rosemount Research and Outreach Center to determine whether known spectral vegetative indices can be used, or new indices developed, to remotely predict alfalfa maturity and corresponding forage quality. A wide range of spectral reflectance (350-2500 nm) was measured in conjunction with destructive sampling, periodically throughout the growth of a stand. From a series of existing spectral indices, REIP (Red Edge Inflection Point) was fit to a quadratic model predicting alfalfa maturity (reported as mean growth stage by weight) (R² = 0.92). A stepwise regression procedure was also used to identify wavebands specific to alfalfa quality, resulting in a linear model to predict RFQ (Relative Forage Quality) (R² = 0.80). Analyses of alternative transformations will continue and expand with measurements from 2015. Additionally, LiDAR (Light Detection and Ranging) measurements are being recorded and analyzed to facilitate yield predictions based on canopy height. Preliminary results indicate a significant linear relationship (R² = 0.84) between LiDAR estimated plant height and actual dry matter yield, up until lodging begins to occur. Both the measurement and analysis of canopy reflectance, and LiDAR based yield predictions have potential to be developed and integrated for viable applications in alfalfa management.