Kass, D.C.L. Tineo B, A.L. Faustino, J. Arriaza, N. Víquez L, E. Rodríguez, M. Jiménez B, J.M.

Performance of Erythrina fusca (Lour.) associated with annual crops following liming of an Acrudoxic Melanudand with 70 percent aluminum saturation - Turrialba (Costa Rica) 1994 - 12 p.

5 tab. 14 ref. Sum. (En, Es)

An experiment was set up in April, 1990 to measure erosion control by alley farming and mulch systems using Erythrina fusca and Inga edulis on a soil with 15-35 percent slope in the CATIE experiment station in Turrialba, Costa Rica. Growth of the first maize crop was extremely poor which was attributed to the high A1 saturation of the soil, which in contrast to mos other soils on the CATIE station was over 70 percent. In October of 1990, 2,6 t ha-1 of a limestone obtained from a quarry less than 15 km from the experimental site was applied to all plots except an uncropped control. Subsequent crop yields were quite satisfactory but tree biomass production remained low. Soil samples taken to an 80 cm depth showed considerable increase of exchangeable Ca to at least 20 cm over the next three years but tree growth remained poor. Nitrate cocentration was higher in the uncropped, unlimed plots especially below thew 80 cm depth where the soil was considered to demonstrate positive charge characteristic. Inoculation and Mg application failed to improve tree biomass production. Crop production reimaned satisfactory over four bean crops with a total application of less than 50 Kg ha-1 of P. Apparently liming has resulted in the mineralization of considerable quantities of organic N and P which for some reason was not available to the trees. Among the hypotheses formulated to axplain the results observed are: 1. Trees absorb the major part their nutrients from soil depths which are as yet not affected by the liming. 2. Liming has produced a soil environment not favorable for tree growth. 3. Erythrina, like several other genera adapted to acid soils is unable to absorb N in the form of nitrates (Foy, 1984) and N fixation may be inhibited by high soil nitrete levels. It was concluded that soil Mg levels were critical to satisfactory E. fusca growth as biomass production was highly correlated with soil and foliar Mg levels. In the fourth year of the experiment, the E. fusca clone though supposedly selected on the same soil as the experiment was replaced with a clone of E. berteroana which produced higher biomass on a trial on the same soil which has been classified as a Acrudoxic Melanudand because of its low bulk density and low levels of exchangeable bases.