Premium
Frequency Distributions of Soil Properties on a Loess‐mantled Missouri Watershed
Author(s) -
Young F. J.,
Hammer R. D.,
Larsen D.
Publication year - 1999
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/sssaj1999.03615995006300010025x
Subject(s) - transect , soil science , pedogenesis , normality , geology , outlier , watershed , environmental science , soil water , hydrology (agriculture) , mathematics , statistics , geotechnical engineering , machine learning , computer science , oceanography
Frequency distributions of soil property values are largely unknown, yet they are important to studies of soil variability. Our objectives in this paper were to (i) evaluate the distributional normality of soil properties within and among upland geomorphic surfaces, and (ii) interpret these distributions in terms of pedogenic factors and processes. Sixty soil properties were measured from 257 pedons along point transects in a 40‐ha upland Missouri pasture. Five of the 60 tested variables approximated normality. About three‐quarters of the distributions were asymmetric (skewed), and over four‐fifths were misshapen (kurtotic). Partitioning the samples by geomorphic surface (ridge, shoulder and backslope) improved the normalities, symmetries and shapes of many, but not all, distributions. A priori assumptions of normality are not warranted for most soil properties. Care should be taken when applying statistical tests that are sensitive to departures from normality. However, asymmetric or misshapen distributions provide meaningful information. Extended tails in skewed distributions represent outliers or inclusions in soil survey map units. These distributions result from variation in depositional environments or the asymmetric effects of pedogenic or hydrologic processes. Kurtotic distributions may indicate the presence within the sample of subpopulations with differing means or variances. Distributions provide important clues to understanding spatial variability.