Premium
Positional, Spatially Correlated and Random Components of Variability in Carbon Dioxide Efllux
Author(s) -
Aiken R. M.,
Jawson M. D.,
Grahammer K.,
Polymenopoulos A. D.
Publication year - 1991
Publication title -
journal of environmental quality
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.888
H-Index - 171
eISSN - 1537-2537
pISSN - 0047-2425
DOI - 10.2134/jeq1991.00472425002000010049x
Subject(s) - spatial variability , homogeneity (statistics) , soil science , carbon dioxide , environmental science , soil respiration , spatial correlation , regression analysis , mathematics , partial least squares regression , soil carbon , statistics , soil water , chemistry , organic chemistry
The ability to distinguish spatial variability either from deterministic trends or from experimental treatment effects contributes to the accuracy and interpretation of field experiments. The objective of this study was to characterize three components of soil and soil plus vegetation CO 2 efllux: the positional trend, spatial correlation, and random variation. Soil CO 2 efflux was measured in a wheat ( Triticum aestivum L.) field at two different dates with differing soil water contents, and soil plus vegetative CO 2 efflux was measured at three grassland sites. Carbon dioxide efflux was determined by the alkali absorption method using static chambers located at 3‐m intervals within a 18 m by 18 m square grid. Positional trends were identified using a multiple regression technique, and with a t ‐test extended into two dimensions. Data were detrended using a generalized least squares (GLS) regression procedure. Semivariograms were used for analysis of spatial correlation and random variability. The assumption of spatial homogeneity (first‐order stationarity) was not founded for CO 2 efflux for four of the five data sets. Positional trends accounted for 16 to 48% of the total variability in these cases. Spatial correlation was not detected, although ignoring positional trends may well have resulted in the opposite conclusion. Spatial structure was affected by the soil water content under wheat. Defining the spatial structure of soil respiration requires determinations under a range of environmental conditions. Advantages of the trend identification and quantification procedures utilized are direct application of common regression techniques, direct evaluation of first‐order stationarity following trend removal and correction for correlated error structure.