No abstract

Soil organic matter (SOM) has been proposed as an index of N supply in paddy soils although field validations are few. We evaluated the relationship between the indigenous N supply (Ni) of the soil-floodwater system and soil organic carbon (SOC) or total N (Nt) in surface soil of long-term fertility experiments (LTFEs) at 11 sites, in 42 farmer's fields with similar soil type, and in the same field in ten consecutive rice (Oryza sativa L.) crops. The Ni was estimated by crop N uptake from plots without applied N (No plots) under otherwise favorable growth conditions. There was a tight linear correlation between yields and N uptake in No plots and tremendous variation in both parameters among LTFE sites, farmer's fields, and in the same field over time. Correlation between Ni and SOC or Nt explained little of this variation. Factors likely to contribute to the poor correlation were: (1) inputs of N from sources other than N mineralization of SOM in surface soil, (2) degree of congruence between soil N supply and crop demand, which is sensitive to soil drying, length of fallow, crop rotation, and residue management, and (3) differences in SOM quality related to intensive cropping in submerged soil. Better understanding of the processes governing the Ni of tropical lowland rice systems would contribute to the development of crop management practices that optimize utilization of indigenous N resources.

SUMMARYNitrogen efficiency from Azolla microphylla or Sesbania rostrata green manure, rice straw, and inorganic fertilizer-N was compared in two long-term experiments with irrigated lowland rice (Oryza sativa L.). Treatments included a control and each nitrogen source alone or in combinations that provided 50% of the total applied nitrogen from an organic and inorganic nitrogen source. All nitrogen sources were applied at equivalent nitrogen rates to 19–22 consecutive rice crops. Residual effects were assessed in two subsequent cropping seasons at one site. Lower grain yield, agronomic efficiency (Δgrain per kg total applied nitrogen), and apparent nitrogen uptake were obtained from green manure and rice straw nitrogen as sole or dual nitrogen sources rather than from a standard split application of prilled urea. Compared to prilled urea, residual effects from green manure or rice straw included a significant increase in soil organic carbon and total nitrogen, and greater extractable soil nitrogen in the vegetative growth period. After panicle initiation there was no residual effect on the rate of crop nitrogen accumulation, and final grain yields were similar regardless of previous nitrogen source. Recycling of rice straw appeared to have greater potential for reducing fertilizer-N requirements than use of green manure because rice straw is often a wasted resource in irrigated rice systems of the humid tropics, the efficiency of rice straw nitrogen in combination with prilled urea is comparable to green manure nitrogen, and the increase in soil nitrogen from rice straw was 50–150% greater than from green manure.

Chemical N fertilizers are usually applied to modern rice ( Oryz.a sativa L.) cultivars, but yield responses are obtained from green manuring as well. To be considered as a viable alternative to chemical N fertilizer in tropical Asia, however, green manures must tolerate the ephemeral waterlogging that precedes rice cultivation. Experiments on an Aerie Tropaqualf soil during three early wet seasons determined green manure N accumulation in 60 d or less, and rice grain yield response to it. In the first experiment (1983), cowpea (Vigna unguiculata (L.) Walp.) and Sesbania cannabina ((Retz.) Poir.) were grown 24, 36, and 48 d before incorporation. Nitrogen accumulation was greatest (79 kg N ha-1 ) in 48-d S. cannabina. Accumulation in cowpea was apparently depressed by waterlogging. Rice grain yield response to green manure N was unaffected by location of green manure cultivation (in situ vs. transported from a nearby field). In the second (1984) and third (1985) experiments, Sesbania sp. were flooded during the last 25 d of growth to assure early wet-season waterlogging. Mean N accumulation in flooded S. rostrata (Brem. and Obrem.), which nodulates on its stem, was 90 kg ha-1 in 48 d and 164 kg ha-1 in 60 d. Nitrogen accumulations in nonflooded S. rostrata were 128 kg ha-1 in 48 d and 198 kg ha-1 in 60 d. Accumulation in a Nepalese S. cannabina cultivar averaged 84% of that in S. rostrata. Rice grain yield response was a function of green manure N regardless of species or age. Both S. rostrata and S. cannabina tolerate waterlogging and accumulate sufficient N in less than 60 d to increase rice yields more than 2 Mg ha-1 • R.R. Furoc and M.A. Dizon, Multiple Cropping Dep., and E.P. Marqueses, Agronomy Dep., Int. Rice Res. Inst. (IRRI),

For many Asian farmers, inorganic N fertilizer for rice (Oryza sativaL.) is costly. Biologically fixed sources of N are being reexamined as alternatives to inorganic N fertilizer. Land that is fallow during the first 5 to 7 weeks of the wet season, before monsoon rains intensify, can be used to accumulate N in a green manure. This experiment determined quantities of N that well‐adapted tropical legumes accumulated in 20 to 45 days and rice yield responses to green manuring with legumes grown for 45 or fewer days. The experiment, repeated in 1981, 1982, 1983, and 1984, was conducted on a fine, mixed, isohyperthermic Aeric Tropaqualf at IRRI, Los Baños, Philippines. Rice was transplanted in factorial arrangements of inorganic N rates (0 and 80 kg N ha 1) and green manure growth durations. In 1981, 1982, and 1983, mungbean [Vigna rudiata(L.) Wilczek] was the green manure, and growth durations were 0 (clean fallow), 20, 30, and 40 days. In 1984, cowpea [Vigna Unguiculata(L.) Walp.] was the green manure, and durations were 0,25,35, and 45 days. Average N accumulation in the oldest green manure was 83 kg N ha 1. Mean rice yield response to 80 kg inorganic N ha 1was 1.0 Mg ha1. Mean response to the oldest green manure was 2.0 Mg ha−1. The study demonstrated that a fast‐growing tropical legume can accumulate more than 80 kg N ha−1in 45 days and that rice yield responses exceeding 2 Mg ha−1are possible from green manure incorporation.