Publications des agents du Cirad

Cirad

Development and validation of a model to describe root length density of maize from root counts on soil profiles

Chopart J.L., Siband P.L.. 1999. Plant and Soil, 214 : p. 61-74.

DOI: 10.1023/A:1004658918388

Root length density (RLD) is an important determinant of crop water and nutrient acquisition, but is difficult to measure in the field. On a soil profile, in-situ counts of root impacts per unit surface on soil profiles (NI) can be used to calculate RLD if crop-specific parameters for preferential root orientation (anisotropy) are known. An improved method for field determinations of RLD was developed and validated for maize at sites in Côte d'lvoire and Burkina Faso. Root anisotropy was measured with cubes of undisturbed soil with 0.1 m sidelength, based on NI observed on three planes oriented perpendicularly to each other. RLD was also measured for the enclosed volume. Repetition of such measurements enabled estimation of the robustness across sites of empirical and geometric models for the relationship between RLD and NI:RLD = NI CO, with CO being the coefficient of root orientation, theoretically equals 2 for an isotropic distribution. Root systems were found to be nearly isotropic, except near the root front (0.3 to 0.5 m), where roots had a preferentially orthotropic orientation. Measured RLD was generally about 50% larger than RLD calculated from observed NI and CO, indicating that at least one of the measurement techniques had a systematic error. The ratio between measured and calculated RLD (CE), which ranged from 0.8 to 2, increased with the age of the plants and decreased with soil depth. CE was therefore introduced as an additional coefficient, resulting in RLD = NI CO CE. The empirical value for CO CE was between 2 and 5. The empirical coefficients CO and CE were the same for the sites in Cote d'lvoire (oxisol with an iron pan at 0.6 to 0.9 m) and Burkina Faso (alfisol with an iron pan at 0.4 to 0.8 m). The model was validated with independent data sets at both sites, and gave satisfactory predictions of RLD on the basis of NI obtained from single soil planes, which can be easily measured in the field.

Mots-clés : zea mays; système racinaire; racine; modèle de simulation; tropisme; croissance; expérimentation au champ; modélisation; côte d'ivoire; burkina faso

Documents associés

Article (a-revue à facteur d'impact)