Biological N2 fixation in wetland rice fields: estimation and contribution to nitrogen balance

This paper 1) reviews improvements and new approaches in methodologies for estimating biological N2 fixation (BNF) in wetland soils, 2) summarizes earlier quantitative estimates and recent data, and 3) discusses the contribution of BNF to N balance in wetland-rice culture. Measuring acetylene reducing activity (ARA) is still the most popular method for assessing BNF in rice fields. Recent studies confirm that ARA measurements present a number of problems that may render quantitative extrapolations questionable. On the other hand, few comparative measures show I ARA's potential as a quantitative estimate. Methods for measuring photodependent and associative ARA in field studies have been standardized, and major progress has been made in sampling procedures. Standardized ARA measurements have shown significant differences in associative N2 fixation among rice varieties. The 15N dilution method is suitable for measuring the percentage of N derived from the atmosphere (% Ndfa) in legumes and rice. In particular, the N dilution technique, using available soil N as control, appears to be a promising method for screening rice varieties for ability to utilize biologically fixed N. Attempts to adapt the 15N dilution method to aquatic N2 fixers Azolla and blue-green algae [BGA]) encountered difficulties due to the rapid change in N enrichment of the water. Differences in natural 15N abundance have been used to show differences among plant organs and species or varieties in rice and Azolla, and to estimate Ndfa by Azolla, but the method appears to be semi-quantitative. Recent pot experiments using stabilized 15N-labelled soil or balances in pots covered with black cloth indicate a contribution of 10-30 kg N ha-1 crop-1 by heterotrophic BNF in flooded planted soil with no or little N fertilizer used. Associative BNF extrapolated from ARA and 15N incorporation range from 1 to 7 kg N ha-1 crop-1 Straw application increases heterotrophic and photodependent BNF. Pot experiments show N gains of 2-4 mg N g-1 straw added at 10 tons ha-1. N2 fixation by BGA has been almost exclusively estimated by ARA and biomass measurements. Estimates by ARA range from a few to 80 kg N ha-1 crop (average 27 kg). Recent extensive measurements show extrapolated values of about 20 kg N ha-1 crop-1 in no-N plots, 8 kg in plots with broadcast urea, and 12 kg in plots with deep-placed urea. Most information on N2 fixed by Azolla and legume green manure comes from N accumulation measurements and determination of % Ndfa. Recent trials in an international network show standing crops of Azolla averaging 30-40 kg N ha-1 and the accumulation of 50-90 kg N ha-1 for two crops of Azolla grown before and after transplanting rice. Estimates of % Ndfa in Azolla by 15N dilution and delta 15N methods range from 51 to 99%. Assuming 50-80% Ndfa in legume green manures, one crop can provide 50-100 kg N ha-1 in 50 days. Few balance studies in microplots or pots report extrapolated N gains of 150-250 kg N ha-1 crop-1. N balances in long-term fertility experiments range from 19 to 98 kg N ha-1 crop-1 (average 50 kg N) in fields with no N fertilizer applied. The problems encountered with ARA and N methods have revived interest in N balance studies in pots. Balances are usually highest in flooded planted pots exposed to light and receiving no N fertilizer; extrapolated values range from 16 to 70 kg N ha-1 crop-1 (average 38 kg N). A compilation of balance experiments with rice soil shows an average balance of about 30 kg N ha-1 crop-1 in soils where no inorganic fertilizer N was applied. Biological N2 fixation by individual systems can be estimated more or less accurately, but total BNF in a rice field has not yet been estimated by measuring simultaneously the activities of the various components in situ. As a result, it is not clear if the activities of the different N2-fixing systems are independent or related. A method to estimate in situ the contribution of N2 fixed to rice nutrition is still not available. Dynamics of BNF during the crop cycle is known for indigenous agents but the pattern of fixed N availability to rice is known only for a few green manure crops.