The potential of legume green manure (GM) as an alternative to mineral N fe
rtilizer in tropical horticulture has received scant attention. The feasibi
lity of meeting N needs of tomato (Lycopersicon esculentum Mill.) with GM w
as studied in six field experiments at three locations in major vegetable g
rowing areas of Taiwan and the Philippines between 1993 and 1995. Legume bi
omass, Nz fixation, and N accumulation, and tomato yield and N uptake were
quantified within a 6-mo experiment cropping pattern. Yields of GM-amended
tomato crops were compared with those amended with fertilizer N (0-150 kg N
ha(-1)). The residual effect of the fertilizing method of a second crop (m
aize; Zea mays L.) was estimated at AVRDC by measures of biomass and N upta
ke 30 d after sowing. Legume N recovery in tomato crops was traced with N-1
5 at Mariano Marcos State University (MMSU), Soybean [Glycine max (L.) Merr
.] harvested at 60 to 74 d accumulated a minimum of 2.8 Mg ha(-1) biomass a
nd 100 kg ha(-1) N in all locations and seasons. A maximum of 6 Mg biomass
ha(-1) and 140 kg N ha(-1) was reached in the wet season (WS) al: AVRDC. In
digofera (Indigofera tinctoria L.) and mungbean [Vigna: radiata (L.) Wilcz.
] biomass yields were more variable and always inferior than soybean yields
. Tomato yields across locations ranged from 3 to 70 Mg fruit ha(-1). Tomat
o yields responded to GM N in the WS in Taiwan and in the northern Philippi
nes, comparing favorably with fertilizer at 38 to 120 kg N ha(-1). No respo
nse to GM N was found in the dry season (DS) at AVRDC or at Bukidnon Resour
ces Company, Inc. (BRCI). The N-15 experiments showed that only a small fra
ction of legume N (9-15%) was recovered by the tomato crop at MMSU. Maize b
iomass and N uptake, following the tomato crop, was increased with soybean
GM compared with the control in the AVRDC WS and DS. Tomato yield response
to GM N is high on infertile soils and tomato N requirement can be substitu
ted fully or partially by GM, depending on soil N mineralization.