Utilization of Spectral Reflectance and Geographic Information System (GIS) Analysis for Bermudagrass Selection and Improvement.

Remote sensing in row crop agriculture has been used to predict vegetative heath and forecast yield for more than two decades.  Little is known regarding the widespread application of this technology for assessing turf bermudagrass (Cynodon [L.] Rich.) genotypes, under various management practices and environmental conditions.  Therefore, a field study was initiated in Tifton, GA to determine how genotype (2 unique hybrids), mowing height (1.3 cm and 3.8 cm), growth regulator (control and Primo), and drought cycle treatments affect spectral reflectance as conveyed through normalized difference (NDVI) and ratio (RVI) vegetation indices.  Spectral reflectance was measured using a Crop Circle ACS 470 (Holland Scientific, Lincoln, NE) mounted to a push-cart 72.5 cm above the ground.  The sensor was equipped to measure light reflectance in the green (550 nm), red (650 nm) and NIR (760 nm). Ground truth consisted of turfgrass color and percent green plot cover measured using digital image analysis, as well as, turfgrass density and quality estimated based on the National Turfgrass Evaluation Program’s (NTEP) visual ratings guidelines using a scale of 1 to 9.  Preliminary results indicate that estimates of turf health using the NDVI and RVI during artificially induced drought cycles correlate very well (r = 0.92, P ≤ 0.0001) with leaf firing and well (r = 0.73, P = 0.0003) with plot color.  These findings suggest that ground-based remotely sensed data can predict traditional turfgrass measurements with less effort in some circumstances, although further testing will be necessary to validate this technology.