Page Banner

United States Department of Agriculture

Agricultural Research Service

Title: Effects of Falling Number Sample Weight on Prediction of Alpha-Amylase Activity

item Finney, Patrick

Submitted to: Cereal Chemistry
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: April 9, 2001
Publication Date: N/A

Interpretive Summary: Field sprouting can damage the value of wheat and flour because of the enzyme, alpha-amylase, which develops in proportion to the amount of field sprouting in the wheat. Thus, field sprouting must be quantified before the wheat is binned by elevator operators and millers. The Falling Number (FN) assay is commonly used to estimate degree of field sprouting in wheat. The official FN method recommends 7 grams sample weight. However, this study definitively showed that a 5 grams sample produced more reliable results by reducing variation between replicates and by improving the FN predictability of alpha-amylase, the enzyme that develops when wheat sprouts in the field when exposed to rainy periods just before harvest. Thus, the FN test is best run with about 5 grams sample, not the 7 grams officially recommended by the AACC. Also, this study showed that a number of flour/water viscosity measurements produced by the Rapid Visco Analyzer more highly predicted alpha-amylase than did the FN. Thus, elevator operators and millers who purchase wheat and must bin the wheat according to the amount of field sprouting, will benefit by this research. Those who now use the FN test can improve the test reliability by reducing sample weight from 7 to 5 grams and for those or others who are interested in an improved method, this study showed that the RVA test is better than the commonly used FN test.

Technical Abstract: As Falling Number (FN) sample weight was decreased from 7, 6 and 5.5 to 5 grams, the coefficient of variation (c.v.) for the test was reduced from 5.75, 2.12 and 1.93 to 1.72%, respectively. Conversely, decreasing (FN) sample weight below 5 grams to 4.5 and 4.0 grams increased the c.v. from 1.72 to 4.27 and 14.47, respectively. The FN test with the greatest reproducibility between sample replicates resulted from the 5 grams sample weight. The regression coefficients (r2) between alpha-amylase activity and FN numbers converted to Liquefaction Numbers ranged from 0.65 to 0.77, for sample, weights between 5 and 7 grams, when considering all 200 samples. When most of the sound samples were not regressed, the LN values regressed by alpha-amylase averaged about r2 = 0.85. By reducing FN sample weight from 7 to 5 grams, a FN value of 350 sec, which is considered essentially sound, was reduced to 215 sec, shortening the test time by 2 minute and 35 seconds. Thus, the FN test is best run with about 5 grams sample, not the 7 grams officially recommended by the AACC. The regression coefficients (r2) between alpha- mylase activity and the converted (1/RVA value) First Peak, Final Peak, Set Back, Trough and Peak Time were nearly all higher than between alpha-amylase and LN values, mostly ranging between r2 = 0.90 and 0.95. In this study, the five RVA traits better predicted alpha-amylase activity than did FN values converted to LN values, at any sample size between 4 and 7 grams.

Last Modified: 4/19/2014
Footer Content Back to Top of Page