USE OF VELVET BEAN TO IMPROVE SOIL FERTILITY AND WEED-CONTROL IN CORNPRODUCTION IN NORTHERN BELIZE

The objective of this research was to mesasure the effect of intercrop ping corn (Zea mays L.) with velvet bean (Mucuna pruriens) oil corn yi elds, soil fertility, and weed control in northern Belize. Two soils w ere used which are widespread in the area. These soils are high in cla y (simectitc) and are well supplied with bases. One soil, the Louisvil le, has a black surface horizon overlying light gray to white calcareo us weathered coral. The other soil was Xaibe which consists of a reddi sh-brown clay over hard Limestone ''cap rock'' at 30-40 cm. Both soils had a relatively high organic carbon (C) and total nitrogen (N) conte nt. We classified the Louisville as Calciustert and the Xaibe as a Ust ropept. The treatments used (carried out annually in 1990, 1991, and 1 992) were residue burn (B), burn plus fertilizer (BF), a velvet bean i ntercrop (G), and a velvet bean intercrop plus fertilizer (GF). Velvet bean intercropping did not have a positive effect on corn yields on t hese soils. For adequate yields to be maintained, it appears that fert ilization with N, phosphorus (P), potassium (K) and certain micronutri ents are required. We did obtain some indication that velvet bean inte rcropping will reduce weed population, including a serious infestation of cogongrass (Imperata cylindrica). Velvet bean did not maintain the soil N content. Corn yields decreased on the Louisville soil at San V ictor from a range of 2,804 kg/ha to 3,181 kg/ha in 1990 to 1,783 kg/h a to 2,065 kg/ha in 1992. There were no significant treatment effects in any year. With the Xaibe soil, no yields were obtained in 1990 beca use of weed competition. In 1991, yields ranged from 555 kg/ha to 1,14 6 kg/ha (no significant treatment effects). In 1992, the fertilized pl ots, BF and GF, had corn yields of 1,391 kg/ha and 1,592 kg/ha, respec tively. The unfertilized plots, B and G, had corn yields of 751 kg/ha and 699 kg/hat respectively. We did obtain sufficient information valu able for persons making fertilizer recommendations on similar types of soil. The Mehlich 3 and the ammonium bicarbonate-DTPA extractant (the latter extractant similar to that used in Belize) did not give good c orrelations between P and K leaf levels and corn yields. Where soil te st P and K were often adequate, corn ear leaf levels were deficient (a nd presumably yields were reduced because of this metabolic deficiency ). We concluded from both soil tests and leaf analyses that the micron utrients copper (Cu), manganese (Mn), and molybdenum (Mo) are seldom, if ever, limiting plant growth and grain yield, although there appears to be seasonal differences. With both soils, soil tests were not good indicators of zinc (Zn) concentrations in the leaf. In 1992, soil tes ts indicated adequacy of this element, although leaf Zn was deficient in all plots on the Louisville soil in 1990, most plots in 1992, and s everal plots on the Xaibe soil in both years.