In East Africa, highland bananas (AAA-EA, Musa spp.) are a primary staple crop for smallholders (<2 ha) who traditionally practice labor-intensive non-mechanized farming, often with common beans (Phaseolus vulgaris) as an understorey intercrop. At the onset of the wet season, farmers till their banana fields to allow the bean intercropping. This study aimed at testing whether tillage with a forked hoe would causes less damage to the banana root system than the traditional hand hoe. Measurements were taken in Walungu / Sud-Kivu in the Eastern DR Congo. A field trial with highland bananas was planted in April 2008 to explore the impact of mulching and tillage on banana performance. It consisted of a randomized complete block design with four treatments and four replicates. On September 2010, each T0 plot (i.e., tilled with blade hoe and mulch removed at onset of wet season) was divided into two sub-plots. The first sub-plot was tilled with conventional hoe (blade) whereas a forked hoe was used in the second sub-plot. Banana root fresh weight and length of cord roots were assessed at 0-10 cm and 10-20 cm soil depths using the core sampling method. No significant difference was found between the two tillage tools, either in terms of root fresh weight or root length. For both blade hoe and forked hoe, tillage decreased significantly the root fresh weight and root length within the 0-10 cm soil layer, but showed little impact at greater depth. We conclude that any type of tillage practiced with a hoe by the farmers in the study area strongly affects the banana root system in the topsoil.

Conservation agriculture techniques (e.g. no-tillage with mulching) have been much promoted as an alternative to conventional tillage in many cropping systems. However, adoption of no-tillage with mulching will only be widely adopted if competitive yields are demonstrated with its use. Our objective was to assess the short-term effect of no-tillage with mulch on the yield of bean (Phaseolus vulgaris L.). Bush bean yield with manual tillage without mulch (T0), no-tillage with banana residues (T1) and no-tillage with Tripascum laxum mulch (T2) were compared during two consecutives growing seasons. Banana residues and T. laxum mulch were applied at the rate of 12.5 tons dry matter ha-1 an-1. A significant (P<0.01) increase in bean yield was found in the first growing season (