PARAMETERS AFFECTING RESIDUE NITROGEN MINERALIZATION IN FLOODED SOILS

If N mineralization from organic residues could be matched to plant up take, N loss will be reduced. However, it is important to understand t he transformations of residues and their controlling factors. Six-week phytotron and two-season field experiments were conducted to determin e the influence of soil properties, residue composition, and soil faun a on N mineralization in flooded soils. Seven legume species, three le gume-rice (Oryza sativa L.) straw combinations, rice straw alone, and an Azolla sp. were incorporated into two soils (an Andaqueptic Haplaqu oll and a Tropofluvent) at 100 mg N kg(-1) dry soil and incubated for 6 wk. Residue N release in clay was about twice that of sandy soil, bu t was not correlated with the residues' N, C/N, water, or polyphenol c ontents. Net N mineralization was correlated to lignin/N (L/N). In the field, Sesbania rostrata Brem. and Oberm. and S. emerus Taub were inc orporated at 60 kg N ha(-1) 1 d before rice transplanting. Additionall y, rice straw was applied at 30 kg N ha(-1) alone or with S. rostrata. Soil exchangeable NH4+-N in unplanted light and dark subplots was mea sured biweekly. As in the phytotron study, initial soil NH4+-N accumul ation rates were higher from residues with relatively low L/N (5. rost rata) than from materials with higher L/N (5. emerus and S. rostrata-r ice straw mixture). Soil invertebrates were absent from dark subplots, where residue N mineralization during the first 6 wk was 30% lower th an in plots exposed to light. The L/N of an applied residue may be a s uitable indicator for predicting its N mineralization rate in hooded s oils. Soil invertebrates may play a role in N mineralization.