Premium
Spatiotemporal Analysis of the Relative Soil Gas Diffusion Coefficient in Two Sandy Soils: Variability Decomposition and Correlations between Sampling Dates at Two Spatial Scales
Author(s) -
Lafond Jonathan A.,
Allaire Suzanne E.,
Dutilleul Pierre,
Pelletier Bernard,
Lange Sébastien F.,
Cambouris Athy.
Publication year - 2011
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/sssaj2010.0419
Subject(s) - spatial variability , environmental science , soil water , spatial ecology , soil science , hydrology (agriculture) , snowmelt , spatial distribution , sampling (signal processing) , coefficient of variation , atmospheric sciences , spatial heterogeneity , tillage , diffusion , geology , agronomy , ecology , mathematics , snow , geomorphology , statistics , remote sensing , geotechnical engineering , filter (signal processing) , computer science , computer vision , biology , physics , thermodynamics
The variability of the relative soil gas diffusion coefficient ( D s / D o ) in space and time is not well known but is important for root respiration, microbial activities, and greenhouse gas emissions. The objectives of this study were to: (i) quantify the spatial variability of D s / D o ; (ii) decompose this variability into components at small vs. large scales relative to the size of the site; and (iii) analyze temporal changes at each spatial scale during the growing season of 2006 and the spring of 2007 at two depths in two sandy soils under potato ( Solanum tuberosum L.)–corn ( Zea mays L.) rotation. The coefficient of variation of predicted D s / D o varied from 30% at 0.15‐m depth to 101% at 0.30‐m depth, possibly because of a change in soil horizons around the 0.30‐m depth. The variability in the data on each sampling date was decomposed into small‐scale (spatial and nonspatial) and large‐scale (spatial) components using the coregionalization analysis with a drift method. Overall, the small‐scale component was predominant at both sites, especially at the 0.15‐m depth. Spatial structures were maintained during 2006 and partially carried over to the next year, particularly at the 0.30‐m depth, except for the large‐scale spatial distribution patterns observed at both depths at Site 1. These results indicate that variability between years may be higher than within a year, probably due to tillage, plant growth, and snowmelt. The study suggests that spatiotemporal variability of D s / D o should be considered in agricultural research and precision farming approaches. Longer term studies would increase understanding of temporal trends in the spatial distribution.