Nitrogen fixation in endophytic and associative symbiosis

N-balance,N-15 isotope dilution and N-15 natural abundance studies provide strong evidence that some tropical grasses, especially sugar cane (Saccharum spp.), wetland rice (Oryza sativa) and kallar grass (Leptochloa fusca) can obtain at least part of their N-needs from biological nitrogen (N-2) fixation. However, these studies have not provided conclusive evidence that these plants are engaged in symbiotic partnerships with any bacteria, as the techniques used do not distinguish between "symbiotic" N-2 fixation on the one hand, and N-2 fixation inputs from free-living heterotrophs and cyanobacteria on the other. Large and diverse populations of heterotrophic diazotrophs can be isolated from the surfaces and rhizospheres of sugar cant, wetland rice and kallar grass, and also from surface-sterilized plants. While the latter suggests that these organisms are living intimately within the plant tissues and hence may be fixing N-2 in symbiotic partnerships with their hosts, "endophytic diazotrophs" have been observed only within the intercellular spaces, vascular tissue, aerenchyma and dead cells of their hosts, not in their living cells. Recent studies have also shown expression by diazotrophs of genes encoding for nitrogenase (e.g., nifA, nifH), in situ expression of the nitrogenase enzyme proteins, and nitrogenase (acetylene reduction) activity in association with their host plants, but have yet to determine whether the bacteria transfer the products of N-2 fixation to the plants. Earlier studies used N-15(2) gas incorporation to show transfer of fixed N in sugar cane, rice, maize (Zea mays), sorghum (Sorghum bicolor) and various forage grasses. However, the long exposure times often used in these studies could mean that the fixed N may have been transferred principally by the death and mineralization of asymbiotic diazotrophs, rather than through direct and rapid transfer as occurs in legume nodules. To demonstrate that N-2-fixing symbioses are present, future N-15(2) experiments should utilize short exposure times (hours), and compare plants inoculated with wild type diazotrophs and nif mutants. Only an integrated and multidisciplinary approach can determine if (a) N-2-fixing symbioses with Gramineae really exist, and (b) if they are effective in agronomic terms. (C) 2000 Elsevier Science B.V. All rights reserved.