NITROGEN-FIXATION BY SOYBEAN IN COMMERCIAL IRRIGATED CROPS OF CENTRALAND SOUTHERN NEW-SOUTH-WALES

Amounts of symbiotic N-2 fixation were monitored using the ureide-sap technique in 36 irrigated soybean crops [Glycine max (L.) Merrill], fr om 26 commercial operations in the Coleambally irrigation area and in the Macquarie Valley of N.S.W. during the summer of 1990-1991. Crop N yields, N-2 fixation, and nodulation were measured twice during crop g rowth and estimates of the seasonal amounts of N-2 fixed were compared with amounts of seed N removed. For each crop, a net N-balance was ca lculated at harvest to determine whether there was a net gain or loss of N in soil as a result of soybean cropping. All crops under study we re well nodulated; however, relative sap ureides collected indicated t hat the proportion of soybean N derived from N-2 fixation (P-fix) rang ed between 21 and 94% at flowering, and 4 and 96% at seed-filling. Sea sonal estimates of P-fix and amounts of N-2 fixed ranged from 52 to 73 % and 103 to 313 kg N ha(-1) at Coleambally, and 13 to 64% and 44 to 2 38 kg N ha(-1) in the Macquarie Valley, respectively. Following seed h arvest these amounts of N-2 fixation resulted in N-balance determinati ons of between -33 to +69 kg N ha(-1) at Coleambally, and -134 to +52 kg N ha(-1) in the Macquarie Valley. The vegetative residues of 11 of the 33 soybean crops for which measurements continued until final harv est were estimated to have contributed more than 15 kg of fixed N ha(- 1) to soil, while N-2 fixation was insufficient to satisfy seed N requ irements in 12 of the remaining crops. The average seasonal P-fix dete rminations were not greatly different at Coleambally between crops con sidered to be in net positive N-balance (65%) and in net negative bala nce (59%), and the final result was more dependent upon soil type, and the average amounts of crop N accumulated and seed N removed than rel iance upon N-2 fixation for growth per se. However, in the Macquarie V alley, average crop N yields and seed N removal were similar across al l soil types, and there was a distinct relationship between average P- fix values and net N-balance (59 cf. 32% for crops in positive and neg ative balance, respectively). Crops with the highest rates of N-2 fixa tion and greatest contributions of fixed N to soil followed several ye ars of cereal cropping, or occurred where soybean had been double-crop ped with wheat. In these situations nitrate concentrations in surface soils were low at sowing. The lowest amounts of N-2 fixation and poore st N-balance arose when the 1990-1991 crop followed 2 or 3 consecutive years of soybean, at sites with a recent history of legume-based past ures, or where N-fertilizer had been applied. Only 4 of the 33 crops f ollowed through to final harvest exhibited evidence of prolonged N-2 f ixation during seed maturation.