Spatial distribution of maize roots by complete 3D soil monolith sampling

The spatial distribution of root length density (RLD) is important for water and nutrient uptake by plants and biomass allocation in the soil. Experimental root assessment is, however, mostly based on methods that encompass only small fractions of the soil volume. The aim of this study was to characterize the three dimensional (3D) spatial distribution of RLD in the soil of a maize crop for plots of 37.5 and 75 cm row spacing. At each plot, a 3D soil monolith of 70 x 40 x 30 (=84,000) cm(3) was completely sampled in form of 84 cubic samples of 10 cm edge length. Roots were washed from the soil and RLD was determined using the line intersect method. In 2004, mean RLD values were 0.41 cm cm(-3) for narrow and 0.34 cm cm(-3) for wide row spacing at row closure (55 days after planting; DAP) and 0.74 cm cm(-3) (1.37 cm cm(-3) in 2003) for narrow and 0.77 cm cm-3 (0.96 cm cm(-3) in 2003) for wide row spacing at tasseling (104 DAP). The CV values for RLD of 48% to 72% in 2004 were first higher for wide than for narrow row spacing but at the later growth stage (tasseling) lower for wide than for narrow. For individual vertical soil slices, CV values for RLD were about 40-60%, irrespective of the orientation of the slice. The results suggest that RLD was related mainly to the spatial location and the plant row structure, and not governed unambiguously by SBD or SWC. The spatially distributed maize root data suggest that variability of RLD parallel to plant rows is not negligible. Any simplified use of 1D or 2D vertical samples at separate locations may lead to erroneous estimations of RLD profiles.