Spatial Variability of Potato Tuber Yield and Plant Nitrogen Uptake Related to Soil Properties. (1st May 2014)

• Record Type: Journal Article
• Authors: Allaire, Suzanne E.; Cambouris, Athyna N.; Lafond, Jonathan A.; Lange, Sébastien F.; Pelletier, Bernard; Dutilleul, Pierre
• Is Part Of: Agronomy Journal. Volume 106:Number 3(2014)
• Page Start: 851
• Page End: 859
• Publication Date: 2014-05-01
• Subjects: Agronomy -- Periodicals; 630
• DOI: 10.2134/agronj13.0468
• Languages: English
• ISSNs: 0002-1962

Abstract

Abstract : Inferring correlations between yield and soil properties is difficult, because the variables involved vary at different scales. The study goal was to describe the spatial distribution of tuber yield (total yield), plant N accumulation (Nuptake ) and soil properties at small and large field scales with a multivariate geostatistical method (Co‐Regionalization Analysis with a Drift), allowing for a correlation analysis of crop and soil properties at the same scale of variability. Two sites with similar growing practices but different pedodiversity were instrumented at 108 sampling points. Total yield was represented mostly by a small‐scale spatial component (<12.4 m) at Site 1, whereas the importance of the small‐scale variability was minor at Site 2. The Nuptake showed a strong spatial structure at small scale at both sites. Large‐scale component of Nuptake was also present at both sites, and correlations with total yield and Nuptake were stronger at this scale. Correlations with soil properties at large scale only indicate a low temperature early in the season and a lack of water later during the season decreased yield whereas lower soil density and higher electrical conductivity increased plant productivity. For high productivity, some irrigation may be needed, even only once during the season. Property measurements should be taken at large scale for precision agriculture by respecting the scale of variability of each property. Small‐scale variability should be used when treatments are of interest, for example, plots no longer than 11 m should be used to compare fertilizer rates at the sites of this study.