Optimizing Color Analysis for Different Treatment Comparisons.

Traditional numeric visual ratings of color are useful for evaluation of turfgrass, but can be highly variable among raters. Alternative methods of color analysis including chlorophyll index (CI), normalized difference vegetation index (NDVI), and digital image analysis have been shown to have much lower variability, while correlating with visual ratings. However, turfgrass color can be affected by both increased tissue chlorophyll content and coating of leaf tissue with substances like iron or pigments, and in these different scenarios, certain methods of color analysis may not correlate with numeric visual ratings. This study investigates the utility of visual numeric ratings, CI, NDVI, and digital image analysis to capture response of turfgrass color resulting from changes in leaf chlorophyll content and leaf coating alterations. Creeping bentgrass, Agrostis stolonifera, plots were treated for three growing seasons on a 14 day interval with ammonium sulfate at 2.4 or 9.8 kg N ha^-1, and half of the plots were treated with ferrous sulfate at 61 kg Fe ha^-1. Treatments were analyzed for color every 14 days with the visual numeric approach, CI, NDVI, and digital image analysis using ImageJ. Visual numeric ratings detected increased color in high N plots, as well as plots receiving Fe. The CI was positively correlated with visual ratings for N treatments, but was negatively correlated for treatments receiving Fe. The NDVI and digital image analysis correlated with visual ratings for both N and Fe treatments. These findings suggest that methods for color analysis should be carefully selected depending on treatment comparisons. While CI may be appropriate for treatments that affect plant chlorophyll content, alternative methods should be utilized for comparison of treatments that produce darker color by coating the leaf blade.