Premium
Elemental Variability Within a Sampling Unit
Author(s) -
Drees L. R.,
Wilding L. P.
Publication year - 1973
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/sssaj1973.03615995003700010028x
Subject(s) - outwash plain , loess , horizon , geology , mineralogy , soil science , loam , sampling (signal processing) , mathematics , soil water , geomorphology , geometry , glacial period , physics , detector , optics
To determine the magnitude of lateral variability within an area of 1 to 2 m 2 , six profiles were sampled at 0.6 m‐intervals at each of three different sites representing loess, till, and outwash deposits. Elemental Ti, Zr, Fe, Ca, and K were determined by X‐ray spectroscopy for 50‐20µ and 20‐5µ fractions of selected horizons within each profile. Coefficients of variation within these deposits ranged from 1 to 13% for Ti, 4 to 20% for Zr, 3 to 20% for Fe, 5 to 13% for Ca, and 2 to 9% for K. The order of variability was loess ≤ till < outwash. No consistent differences in elemental variability were observed between size fractions or horizons (i.e. A, B, or C). Elemental data from a single horizon within a vertical profile will seldom permit mean estimates closer than ±10 to 20% (95% confidence interval) for loess and till deposits, and ±20 to 45% for the more variable outwash deposits. Three lateral subsamples within a given horizon would permit mean estimates of above elements within ±10% in loess and till but 20 subsamples would be required to achieve similar accuracies in some outwash deposits. Such data emphasize the need to multiple subsample horizons in a lateral dimension if mean estimates are to be accurate within 10 to 20% over an area approximating a pedon or sampling pit. The number of subsamples for a given horizon is a function of the element under consideration, nature of deposit, and precision of mean estimate desired.