Digital Image Analysis Method for Estimation of Fusarium‐Damaged Kernels in Wheat. Issue 5 (1st September 2014)

• Record Type: Journal Article
• Authors: Maloney, Peter V.; Petersen, Stine; Navarro, Rene A.; Marshall, David; McKendry, Anne L.; Costa, Jose M.; Murphy, J. Paul
• Is Part Of: Crop science. Volume 54:Issue 5(2014)
• Page Start: 2077
• Page End: 2083
• Publication Date: 2014-09-01
• Subjects: Crop science -- Periodicals; Cultures -- Périodiques; Cultures de plein champ -- Périodiques; Crop science; Nutzpflanzen; Zeitschrift; Pflanzenbau; Periodicals; 633
• DOI: 10.2135/cropsci2013.07.0432
• Languages: English
• ISSNs: 0011-183X

Abstract

ABSTRACT: Fusarium head blight (FHB) of wheat ( Triticum aestivum L.), or head scab, causes a reduction in grain yield and quality through the formation of shriveled, dull‐grey seeds called "tombstones" or Fusarium‐damaged kernels (FDK). Fusarium‐damaged kernels are commonly quantified on a percentage basis by visually separating damaged from healthy kernels following harvest, in a process that is both time consuming and labor intensive. The objective of this study was to evaluate an alternative method for quantifying FDK through the use of the digital image analysis program ImageJ. The 'NC‐Neuse' × 'AGS 2000' F5 –derived recombinant inbred population of 172 lines and the NC‐Neuse ×'Bess' double haploid population of 112 lines were used in this study. NC‐Neuse and Bess were moderately resistant and AGS 2000 was susceptible to FHB. The populations were evaluated under moderate to heavy FHB epidemics in a total of five environments in North Carolina, Maryland, and Missouri with two to three replications per environment during the 2010 to 2011 and 2011 to 2012 seasons. Following hand harvest and grain processing, FDK was estimated by (i) visual separation of diseased kernels and (ii) digital image analysis using ImageJ on captured images of grain samples. The correlation between the two methods ranged from 0.72 to 0.80 over five environments. A lower correlation was observed in one environment due to cracked and broken kernels in the samples. Digital image analysis was three times faster than the visual method and estimated FDK on a larger scale per plot sample, whereas labor and time constraints limited the sample size for the visual method. Digital image analysis was consistent over samples and appears well suited as an alternative form to estimate percent of FDK in grain that is not damaged in other ways.