Sample size and heterogeneity effects on the analysis of whole soybean seed using near infrared spectroscopy

Seth L. Naeve, Rob A. Proulx, Brent S. Hulke, Tracy A. O'Neill

Research output: Contribution to journalArticlepeer-review

4 Scopus citations

Abstract

Soybean [Glycine max (L.) Merr.] breeders and physiologists are commonly interested in evaluating single plants for seed protein and oil concentration. Moreover, breeders tend to prefer nondestructive techniques so that soybean seed may be retained for variety development. Near infrared spectroscopy (NIRS) technology can be an inexpensive and convenient method for nondestructive analysis of seed; however, many instruments require a 300 to 500-g sample size to operate. Recently, Perten Instruments, Inc. developed a 20 g "breeder's" cup for use with their DA 7200 diode array NIRS instrument. This study was initiated to explore the accuracy and repeatability of this very small cup relative to the standard 300 g cup. While no significant protein bias was discovered, a positive 4 g kg-1 bias for oil prediction was found. Seven to eight repeated scans of the breeder's cup were required to increase cup precision compared to a single scan using the standard large cup, potentially due to the large amount of protein and oil heterogeneity within examined seed lots. While the precision of the protein and oil estimation of both cup sizes increased with repeated scans, the low precision noted with the breeder's cup virtually mandates multiple scans or destructive sampling through grinding.

Original languageEnglish (US)
Pages (from-to)231-234
Number of pages4
JournalAgronomy Journal
Volume100
Issue number1
DOIs
StatePublished - Jan 2008

Fingerprint

Dive into the research topics of 'Sample size and heterogeneity effects on the analysis of whole soybean seed using near infrared spectroscopy'. Together they form a unique fingerprint.

Cite this