Tillage and varietal impacts on tomato (Solanum lycopersicum L.) production on an ultisol in central Cameroon

Across Africa the production of high value horticultural crops is being recognized as a possible route out of rural poverty. To determine at which level of intensification tomato yields were maximized and which tomato cultivar responds best to intensification, an on-farm, factorial trial was conducted at Essong Mintsang in the central region of Cameroon on a rhodic Kandiudult. Current farmer practice of manual tillage yet not destumped was compared with either reduced input (no tillage, not destumped) or increased input (no tillage yet destumped, manual tillage and destumped, mechanical tillage and destumped). Yields and yield components of three tomato varieties were determined to assess if changes in intensity of land preparation can improve soil physical properties and thus yields. At 6 weeks after planting, cultivar Rossol had more flowers (521,126 ha(-1), p = 0.09) than cv. Rio Grande (403,930 ha(-1)), with cv. Roma (504,718 ha(-1)) not being significantly different from either variety. At harvest, across land preparations the cv. Rossol produced higher yields (8.12 Mg ha(-1), p = 0.005) than cv. Roma (6.05 Mg ha(-1)) and cv. Rio Grande (4.46 Mg ha(-1)); the latter being significantly different (p = 0.03) from both others. Tomato total and marketable yields were significantly higher in the destumped tractor till, destumped manual till and stumps-retained manual till treatments than in the stumps retained notill treatment. Yields in destumped no-till treatment were not significantly different from any of the treatments. Total fresh yields of cvs Roma and Rossol increased when the soil was tilled, while cv. Rio Grande had no response to land preparation. Tomato pest and disease damages were not affected by land preparation. Weeding took less time (p = 0.05) in destumped manual till (153 h ha(-1)) and tractor till (157 h ha(-1)) than in not destumped manual till plots (243 h ha(-1)). Soil aggregates were least stable in the destumped, tractor till treatment, with significantly lower MWD (p = 0.02), and higher mesoaggregate proportions (p = 0.05) than in the other treatments. Across tomato cultivars and treatments, the marketable fruit yield could be predicted by clay, macroaggregates and bulk density in the 0-10 cm layer: marketable yield = 0.12 clay -0.14 macroaggregate + 12.20 (with r(2) = 0.50, n = 15, p = 0.016) and for the 10-20 cm soil layer: marketable yield = 0.17 clay + 5.15 bulk density - 10.25 (with r(2) = 0.58, n = 15, p = 0.005). Early flowering and fruit production combined with nematode resistance were probably the main contributing factors to the high yields of cv. Rossol. (C) 2012 Elsevier B.V. All rights reserved.