Jk. Daun et S. Symons, How green is green? Sampling and perception in assessing green seeds and chlorophyll in canola, J AM OIL CH, 77(11), 2000, pp. 1209-1213
Green seeds are used as a grading factor in estimating chlorophyll in canol
a and rapeseed in the Canadian and U.S.A. grading systems. This work examin
es the effect that sampling and perception have on the estimation of green
seeds as well as the effect that sampling has on the determination of chlor
ophyll. individual seed analysis indicated that in order to be considered a
s green, seeds needed to contain between 200 and 400 mg/kg chlorophyll. Var
iation due to binomial sampling played a predominant role in the error in d
etermining the green seed levels in canola. Sampling of large numbers of se
eds, as in the loading of export shipments, reduced the error. Binomial sam
pling also contributed to the error in chlorophyll determination even with
sample sizes as large as 500. Differences in perception of green also were
noted between individuals with coefficients of variation as high as 50% at
the 1% green seed level. The combination of perception error and sampling e
rror may result in samples of 1,000 seeds drawn from a mass with 2% green s
eeds having green seed counts ranging from 0.96 to 3.04%, 19 times out of 2
0.