ISRR 2018

Spatial Root Distribution in a Winter Wheat/Winter Faba Bean Intercropping System using FTIR Spectroscopy for Root Species Identification

author.DisplayName author.DisplayName author.DisplayName
Department for Crop Sciences, Division of Agronomy, University of Goettingen, Germany

The spatial root distribution and root morphology of a species is generally altered by intra- and interspecific competition. The major challenge of studying belowground interactions of multispecies ecosystems is to identify roots on a species level. Even though visual species identification has been successful for some species (e.g. Fagus s., Quercus r.), this process is time consuming and has a high error rate for species with a similar root structure (e.g. grass) or young roots. We used Fourier transform infrared (FTIR) spectroscopy for characterization of species root proportions in a winter faba bean/winter wheat intercropping system to analyze their species specific root distribution. Roots of pure stands and 50%/50% substitutive row intercrops were sampled by root auger on the sown rows and between rows of each species in a two year field trial near Goettingen (Germany). Dry and ground roots of pure stands were used to prepare samples of bean and wheat root mixtures with known composition to calibrate and validate a FTIR species prediction model. Root absorption spectra showed species specific peaks and the species were clearly separated by the cluster analysis. We found out that wheat and bean roots spread into the neighboring species row and accounted there on average for 15 % of the total root biomass. Wheat in pure stands had larger root proportions at deeper soil layers whereas the opposite was true in mixtures. We conclude that the interspecific competition with bean affected the root distribution of wheat which indicates a possible root partitioning of wheat. With our study, we demonstrated that FTIR spectroscopy is a suitable tool for root species identification. A measuring time of less than 30 seconds per sample makes it to promising method for high throughput investigations of spatial root distribution of different species.









Powered by Eventact EMS