In Central America, the traditional cropping system milpa de ano (summer season) and tornamilpa (winter season) were compared over 3 yr (2007-2009). Our experimental objectives were to measure the performance of a maize ( Zea mays L.)-velvet bean [ Mucuna pruriens (L.) DC. subsp. utilis (Wight) Burck] milpa system throughout the summer and winter cultivation, to detect any problems associated with velvet bean use, and to determine the contribution of this tropical legume to soil fertility and maize productivity. In each crop season (separated in space and time) we used a completely randomized design with a 2*2 factorial arrangement of treatments with five repetitions each: without velvet bean and without fertilizer (-V-F), with velvet bean and without fertilizer (+V-F), without velvet bean and with fertilizer (-V+F), and with velvet bean and with fertilizer (+V+F). Results showed that in the winter milpas the presence of velvet bean significantly increased the soil pH, organic matter content, total N, and decreased soil bulk density. In both cycles (winter and summer), treatment with velvet bean (+V) produced higher grain yield, while the treatment without velvet bean (-V) had a lower production. We concluded that the use of velvet bean in the winter milpa contributed to the restoration of soil fertility and increased yield maize in agricultural systems of the small-holder farmers based on low external input.
