Plant Colorants Interfere with Reflectance‐Based Vegetation Indices

Abstract

Reflectance‐based vegetation indices are commonly used to quantify turfgrass color, estimate chlorophyll content, and make inferences about plant health. These methods are often substituted for visual color ratings, as they are well correlated but more objective. This study examines the utility of reflectance‐based vegetation indices for evaluating turfgrass treated with iron or plant colorants. In a study in Madison, WI, creeping bentgrass (Agrostis stolonifera L.) was treated with ammonium sulfate (2.4 or 9.8 kg nitrogen [N] ha−1), Turf Screen (untreated or 4.0 L ha−1), or ferrous sulfate (untreated or 12.2 kg Fe ha−1). Visual color ratings, chlorophyll index (CI), normalized difference vegetation index, normalized difference red edge, red, and near‐infrared (NIR) reflectance were able to detect differences in N rate. When colorant or iron was applied, however, reflectance methods could no longer detect these differences in N rate. In a study in Ithaca, NY, the application of four colorants increased visual color ratings. The pigments Harmonizer and Foursome also increased CI, whereas Turf Screen and Green Lawnger did not. Three of the four pigments decreased the chlorophyll content in leaf tissue from 1 July to 22 August, despite having greater or equal CI values. Spectral analysis of six colorants revealed that these products reflect light in the red and NIR regions and likely interfere with reflectance‐based vegetation indices, which measure leaf reflectance in these same regions of the spectrum. Overall, these findings suggest that reflectance‐based vegetation indices should not be used to make inferences about plant health in studies where a colorant or iron is applied.

Influence of synthetic phthalocyanine pigments on light reflectance of creeping bentgrass

“…These data are congruent with recommendations by Obear et al. (2017) but further explain specific regions of pigment interference within the far‐red and NIR spectra. Reynolds et al.…”

Durability and concentration of organic colorant in the visual quality of Discovery TM bermudagrass

“…However, all results were far below the value considered ideal for bermudagrass (445) proposed by Godoy et al (2012). Obear et al (2017) also found no statistical difference in their experiment on the use of colorant, observing values of 221 and 205 in treatments with and without colorant, respectively, which are higher than the results of the present study (Table 2).…”

“…With respect to durability of the colorant, only the reflectance index accurately captured the data, where results at 10 days already showing a degradation of the product (Table 2). This result is similar to that described by Obear et al (2017) in an experiment with the use of colorant in bentgrass, where an interval of 14 days was used. However, Braun et al (2016) observed good results in the application of the same colorant used herein on the visual quality of Zosya japonica 'Chisholm'.…”

“…The results are similar to those reported, who observed a negative effect on the evaluation of NDVI in bentgrass (Agrostis stolonifera) after colorant application. According to Obear et al (2017), NDVI is not an accurate tool to assess turfgrass subjected to colorant application, since the reflectance of the pigment interferes in the device's ability to accurately collect a reading. On the other hand, there were no significant differences for NDVI in relation to time, but with a noticeable reduction in values throughout the study (at 10, 20 and 30 DAA).…”

Guiding fall fertilization of cool‐season turfgrass lawns with NDVI sufficiency index

“…Although the conversion of raw NDVI values to SI will mitigate many problems that affect reflectance, there are still several problems associated with NDVI that could reduce its effectiveness as a guide for N fertilization. A common practice in turfgrass management is to add a colorant or chelated iron to increase the color quality without additional N. On a creeping bentgrass ( Agrostis stolonifera L.) putting green, NDVI and other reflectance conversions were unable to determine differences in N rates when colorants or iron were applied (Obear et al., 2017). It was concluded that these products reflect light in the red and near‐infrared regions and likely interfere with reflectance‐based vegetation indices that measure canopy reflectance in the same regions of the light spectrum.…”