Effects of tillage and mulching on runoff under banana (Musa spp.) on a tropical Andosol P. Cattan a, * , Y.-M. Cabidoche b , J.-G. Lacas c,d , M. Voltz c a CIRAD, FLHOR Department, Station de Neufcha ˆteau, Sainte-Marie, 97130 Capesterre Belle Eau, France b INRA, UR Agrope ´doclimatique de la Zone Caraı ¨be, Domaine Duclos, 97170 Petit-Bourg, France c INRA, UMR LISAH, Bat. 24, 2 place Viala, 34060 Montpellier Cedex 1, France d CEMAGREF, UR Qualite ´ des Eaux et Pre ´vention des Pollutions, 3 bis quai Chauveau, CP220, 69336 Lyon Cedex 9, France Received 13 May 2004; received in revised form 25 January 2005; accepted 3 February 2005 Abstract This article assesses the extent of runoff in banana (Musa spp.) plantations on an Andosol of high water infiltration capacity (saturated hydraulic conductivity, K sat , of over 60 mm/h) with high rainfall levels (2500–4500 mm of rain per year). This is an important issue, since the large quantities of inputs applied to banana plantations are likely to be displaced by runoff. Runoff was assessed over two crop cycles for two crop management strategies, viz. (1) tillage followed by furrowing and planting, and (2) planting in holes without prior tillage, with different soil cover management techniques for the second crop, viz. (1) mulching with harvest residues in every other interrow and (2) mulching with harvest residues in every interrow, whereas the soil was left bare for the first cycle. These strategies correspond to current practice among farmers. The results showed that runoff was generally moderate on the Andosol: runoff amounted to less than 55 mm in the first cycle with a total of 771 mm of rainfall, and about 70 mm in the second cycle, with a total of 1232 mm of rainfall. The mean runoff coefficients (RC) per rainfall event ranged from 5 to 11%. The differences in runoff between these two soil management strategies were particularly marked in the first cycle with no soil cover. Runoff was up to five times higher in the untilled plot (RC maximum value of 12%) for light rainfall events (less than 5 mm), notably due to the lower infiltration capacity of the soil (K sat of 75 mm/h in the untilled plot to 265 mm/h in the tilled one) and despite the indirect hydraulic connectivity with the outlet, which increased the water residence time in the plot, compared to the tilled plot. The differences were reversed in cycle 2 for light rainfall events, where runoff in the tilled plot (RC maximum value of 8%) reached 10 times that in the untilled plot,as a result ofmulching in the untilled plot,while in the tilled plot, the partially mulchedsoil reconsolidated.However, the runoff coefficient did not depend on the soil surface condition in the event of heavy rain (more than 10 mm), when runoff was widespread. Then, the maximum runoff coefficient per rainfall event never exceeded 34%. The results also showed that to compare strategies and soil management methods, it is crucial to analyse phenomena over successive crop cycles. In our case, this revealed that a low-mechanized strategy should not be ruled out, in view of the degree of runoff observed in the first cycle. # 2005 Elsevier B.V. All rights reserved. Keywords: Runoff; Andosol; Banana; Tillage; Mulching; Wheel tracks; Rainfall intensity www.elsevier.com/locate/still Soil & Tillage Research 86 (2006) 38–51 * Corresponding author. Tel.: +33 590 86 17 74; fax: +33 590 86 80 77. E-mail address: philippe.cattan@cirad.fr (P. Cattan). 0167-1987/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.still.2005.02.002