Root orientation of four sorghum cultivars: application to estimate root length density from root counts in soil profiles

Abstract

____________________________________________________________________________________________________________________________________________________________________________Abstract: The root length density (RLD) distribution in the soil is a key factor for water and nutrient uptake but its direct assessment is not easy. A field method is presented here to assess RLD of sorghum (Sorghum bicolor L. Moench) from the root intersection density (RID) and taking root orientations into account. The method was developed with four varieties cropped in 2006 in Bambey (Senegal) on a sandy soil, and validated on the basis of independent data obtained on two other varieties cropped in 2008 in Thiès in Senegal on a sandy loam soil. Sets of 1 × 10 -3 m 3 cubic undisturbed soil samples were collected at different positions and at three different dates for the four varieties during the first experiment. RIDs of thick and fine roots were assessed on the sides of the soil cubes and RLDs were measured in the volumes. Fine roots appeared to be isotropic when thick roots were horizontal near the surface and gradually became vertical in deeper horizons. RLD was tenfold lower for thick roots than for fine ones. This led to the development of a model to determine RLD from RID. The relationship for all roots was modeled as RLD = RID × CO, where the root orientation coefficient (CO) is equal to 1.97, indicating that the global root distribution is almost isotropic (CO = 2 for isotropic conditions). The model was validated on the basis of independent data from the second experiment. This model enables RLD mapping and description of RLD variability on sorghum from RID observations on trench profiles (with 10 × 10 cm or 5 × 5 cm surface area used for RID measurement).

Sugarcane root length density and distribution from root intersection counting on a trench-profile

“…However, according to theoretical models (Melhuish and Lang, 1968), it was possible to estimate RLD from the root intersection counting and models were validated for maize (Chopart and Siband, 1999), sorghum (Chopart et al, 2008b) and rice (Dusserre et al, 2009). It would now be of interest to assess this root mapping methodology with these crops, especially those with high spatial variability in their root systems.…”

Towards an empirical relationship between root length density and root number in windbreak-grown cadaghi (Corymbia torelliana) trees

“…Root orientation varied by root size in sugarcane (Saccharum spp.) (Chopart et al 2008a) and sorghum (Sorghum bicolor) (Chopart et al 2008b); larger roots (>1 mm diameter) showed horizontal growth near the surface but vertical growth in deeper soil. Fine secondary roots (<1 mm in diameter) were isotropic.…”

“…Fine secondary roots (<1 mm in diameter) were isotropic. Sorghum roots were isotropic when roots of all diameters were pooled together (Chopart et al 2008b). This indicates that although roots in some diameter classes may actually be anisotropic, they may appear isotropic when pooled data (of all diameter classes) is used.…”

“…Measuring RLD takes about five times the time required to measure root number (N) (Vespraskas and Hoyt 1988). Root number is a good predictor of root length in some species (Chopart and Siband 1999;Chopart et al 2008b;Lopez-Zamora et al 2002), and a relationship has been developed to indirectly estimate RLD from N. When roots are randomly distributed in three dimensions (isotropic) in a small volume of soil, geometric theory has proposed a RLD=2N relationship (Kendall and Moran 1963;Melhuish and Lang 1968). When roots are anisotropic, the average number of roots in three-dimensions (N AVG ) can be used (Baldwin et al 1971).…”

“…When roots are anisotropic, the average number of roots in three-dimensions (N AVG ) can be used (Baldwin et al 1971). This relationship has been tested in several species with varying results (Adegbidi et al 2004;Bennie et al 1987;Bland 1989;Chopart et al 2008aChopart et al , 2008bEscamilla et al 1991;Lopez-Zamora et al 2002).…”