Relationship of multispectral radiometry data to qualitative data in turfgrass research

Citation
Le. Trenholm et al., Relationship of multispectral radiometry data to qualitative data in turfgrass research, CROP SCI, 39(3), 1999, pp. 763-769
Citations number
24
Categorie Soggetti
Agriculture/Agronomy
Journal title
CROP SCIENCE
ISSN journal
0011183X → ACNP
Volume
39
Issue
3
Year of publication
1999
Pages
763 - 769
Database
ISI
SICI code
0011-183X(199905/06)39:3<763:ROMRDT>2.0.ZU;2-X
Abstract
Multispectral radiometry provides a method for assessing plant light reflec tance in the visible and near-infrared ranges. Reflectance of narrow wavele ngth ranges as well as ratios of different ranges have been highly correlat ed with absorbency of photosynthetically active radiation, leaf area index (LAI), and plant response to stresses. The objectives of this research were to determine if data obtained by multispectral radiometry might accurately correlate with qualitative data (used as rapid estimates of color, density , and uniformity) typically used in turfgrass research. Furthermore, it was our objective to determine whether multispectral radiometry could discrimi nate between stressed and nonstressed turfgrass plots. This research was co nducted in two consecutive trials during 1997 on seven seashore paspalum (P aspalum vaginatum Swartz) ecotypes and three hybrid bermudagrass (Cynodon d actylon L. x C. transvaalensis Burtt-Davy) cultivars established on a nativ e Appling (clayey, kaolinitic, thermic Typic Kanhapludult) soil at the Univ ersity of Georgia Experiment Station in Griffin, GA. Reflectance at 661 and 813 nn, as well as the ratios normalized difference vegetation index (NDVI , computed as R-935 - R-661/R-935 + R-661), infrared/red (IR/R) (LAI, compu ted as R-935/R-661), Stress1 (R-706/R-760), and Stress2 (R-706/R-813) were highly correlated with visual turf quality, shoot density, and shoot tissue injury (STI) ratings, but not with shoot clipping growth. Regression analy sis also indicated very strong associations with each of these qualitative variables, but not with shoot growth data. Additionally, spectral measureme nts at all wavelengths except 559 and 706 nm were able to consistently dist inguish between wear-treated and untreated plots; This methodology may have applications in both turfgrass research and management to provide quantita tive measures with physiological significance to traditional visual qualita tive estimates of shoot aspects.