Geostatistical Analysis and Scaling of Grapevine Root Distribution

  • A.W. Linsenmeier Forschungsanstalt Geisenheim, Fachgebiet Bodenkunde und Pflanzenernährung, von Lade Strasse 1, D-65366 Geisenheim, Germany
  • O. Löhnertz Forschungsanstalt Geisenheim, Fachgebiet Bodenkunde und Pflanzenernährung, von Lade Strasse 1, D-65366 Geisenheim, Germany
  • R. Lehnart Forschungsanstalt Geisenheim, Fachgebiet Bodenkunde und Pflanzenernährung, von Lade Strasse 1, D-65366 Geisenheim, Germany

Abstract

Geostatistical analysis was conducted for the root distribution of Vitis berlandieri x Vitis riparia using dispersion
index, fractal dimension, autocorrelation and semivariance. The data were derived from the observation of roots
of six 12-year old Riesling/5C vine plants in a field experiment using minirhizothron technique. The dispersion
index (DI) indicated the clustering of roots. Autocorrelation as a function of the distance lag showed that a higher
DI was related to higher autocorrelation at small lags. Small scale (< 3cm) spatial analysis using variograms,
showed a clustering of roots at short distances (< 6cm), but also a periodicity at greater distances (16cm) with hole
effects in the variograms. The spatial variance for small scale was 60-85% within a range of 5-8cm. At medium
scale (5-10cm) the spatial variance decreased to 0-20%. Geostatistical analysis is a useful tool to demonstrate
variation in root distribution at plant level and to improve root sampling. Although the different geostatistical
tools were related, it was not possible to deduce one result from the other quantitatively.

Downloads

Download data is not yet available.
Published
2011-12-07
Section
Articles