Premium
A Comparison of Five Methods for Expressing Aggregation Data
Author(s) -
Schaller F. W.,
Stockinger K. R.
Publication year - 1953
Publication title -
soil science society of america journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.836
H-Index - 168
eISSN - 1435-0661
pISSN - 0361-5995
DOI - 10.2136/sssaj1953.03615995001700040002x
Subject(s) - geometric mean , mathematics , fraction (chemistry) , sieve (category theory) , statistics , sample size determination , correlation , soil science , chromatography , analytical chemistry (journal) , chemistry , combinatorics , geometry , geology
Five ways of expressing aggregation data have been in common use in recent years. These are: percent aggregates greater than 2 mm, 1 mm and 0.25 mm, and geometric mean and mean weight‐diameter. To determine the advantages or disadvantages of the several indices, aggregation results obtained by the Yoder method on several hundred soil samples from three locations in Iowa have been compared. Correlation coefficients were obtained between the mean weight‐diameter and the percent of aggregates > 2 mm, > 1 mm, and > 0.25 mm, and also the geometric mean. The geometric mean was also correlated with the three size separates. The results indicate that a single size fraction such as the > 2 mm or > 1 mm can be satisfactorily used to express soil aggregation. This method of expression would be as reliable (to within the 1% level of significance) as that indicated by the two indices, mean weight‐diameter and geometric mean, which utilize all size fractions. However, it was shown that more replications are necessary for equal accuracy when only a single size fraction is used. If time is to be saved in making the laboratory determinations, it is suggested that a size fraction such as the > 2 mm or > 1 mm might be determined more quickly and with greater accuracy by developing a method using one or two sieves and a larger soil sample. The final selection of an index for expressing soil aggregation must be based on the ability of the index to correlate with crop response.