Potential use of dredged material to improve marginal agricultural soils: I. Physical properties and growth

Improving marginal or degraded land for agricultural or recreational uses with dredged material is a possible alternative to the present practice of land disposal. The purposes of this study were (i) to determine the physical properties of dredged material, marginal soils, and mixtures of the two; and (ii) to compare dry matter production of crops grown in the greenhouse on these materials. Samples of dredged material and marginal soils were collected from 10 locations in the eastern and central United States. In addition, three productive soils from Minnesota were chosen as reference materials for comparison. The texture of the dredged material samples ranged from sand to clay. The marginal soils were chosen for their extreme textural differences from the dredged material. The bulk density of the medium- to fine-textured dredged material samples was low, ranging from 0.74 to 1.21 g/cm³. The bulk density of the two coarse-textured dredged material samples was similar to that of the coarse-textured marginal soils. The addition of fine-textured material to coarse-textured material resulted in intermediate bulk densities. The water-holding capacity of the fine-textured dredged material samples was extremely high as compared with that of fine-textured productive soils from Minnesota. The two coarse-textured dredged material samples had very low water retention values. The 0.33- and 15-bar percentages for the mixtures of dredged material and marginal soil were intermediate between the values of the samples of dredged material and marginal soil alone. Plant yields in the greenhouse were greater from the eight fine-textured dredged material samples than from the marginal soils. When dredged material was mixed with marginal soil, plant yields were intermediate betwen yields for either dredged material or marginal soil. When location effects were taken out, an F-test on multiple regression analysis showed that percent sand and silt were significant (p = 0.01) in explaining the yield differences. These regression equations are used to predict a range of sand and silt content at which the dry matter production is near maximum.