On-farm evaluation of ridging and residue management options in a Sahelian millet-cowpea intercrop. 2. Crop development

In the Sahel, promising technologies for agricultural intensification include millet stover mulching and ridging. A four year on-farm experiment was set-up in order to assess the effect of various combinations of these two technologies on crop development and yield in a millet (Pennisetum glaucum (L.) R. Br.) - cowpea (Vigna unguiculata (L.) Walp.) intercropping system. Treatments included bare surface, ridging, a surface applied banded millet stover mulch (2 t ha(-1)) and a banded millet stover mulch (2 t ha(-1)) buried in ridges. The latter three treatments were implemented exclusively in the cowpea rows, with an annual rotation between the millet and cowpea rows. On bare and ridged plots, millet yields fell below 100 kg grain ha-1 after the first year. This was ascribed mainly to soil acidification and loss of soil organic, matter rather than to soil physical constraints or water availability despite extensive surface crusting and high soil penetration resistance and bulk density. Compared to the bare plots, ridging increased cowpea hay production by 330% over the four years which was attributed to lower soil penetration resistance and bulk density but also to a reduction of 0.15 cmol(+)kg(-1) exchangeable acidity in the ridges. Except during the severe drought year of 1997, millet grain yield in the banded mulch treatment remained fairly stable over time at 526 +/- 9 kg ha(-1). However, a detailed analysis revealed yield compensation mechanisms between various yield components depending on the timing of occurrence of the abiotic stresses. Cowpea productivity was always higher in buried banded mulch plots than in surface applied banded mulch plots but the former treatment appeared unable to sustain millet yields. This decline was attributed to a greater nutrient uptake by cowpea and more rapid acidification in the buried mulch treatment compared to the banded mulch treatment