Observations have been made which suggest that clover is more effective than ryegrass in developingrapid improvements in soit structure. To test this hypothesis, grass, grass/clover and clover swards were grown in undisturbed field cores. Significant differences in drainage rates and plant biomass production, and visible changes in soil structure, all supported the hypothesis that clover improves structure rapidly, but there were no significant differences in bulk density, porosity or aggregate stability between the treatments. Soil moisture characteristic curves of soil/peat/grit compost planted with grass, grass/clover or clover and grown for two years indicate that a more free-draining structure develops under clover than under grass due to a higher ratio of macro-to micropores.

The use of single random primers, selected in the absence of target sequence information, has been shown to be effective in producing DNA amplifications that provide fingerprints which are unique to individual organisms. DNA amplification by random priming was applied to the DNA from isolates of Rhizobium leguminosarum biovar trifolii. Amplification products were produced using a number of primers, and the resulting fingerprints allowed strain differentiation. However, the effectiveness of primers was dependent upon length and GC content. It was also possible to amplify DNA directly from cells in culture and in nodule tissue. Lysis of these cells was achieved simply through heat applied in the initial DNA denaturation stage of the thermal reaction. The ability to produce varied amplification patterns from different Rhizobium isolates, especially directly from nodules, gives this method potential for use in examining genetic structures and relationships in Rhizobium populations.

Vicia faba, field bean . Rhizobium leguminosarum . symbiotic variation . SUMMARY Six Vicia faba populations were grown in all possible combinations with six Rhizobium leguminosarum strains in order to estimate the relative importance of the three genetic components of symbiotic variability . Additive genetic effects of host genotype and rhizobium genotype accounted for only 8 .9°< and 11 .8% of the total phenotypic variation . Non-additive variation attributable to specific host genotype x rhizobium genotype interactions was by far the largest component of variation and accounted for 73 .8% of phenotypic differences . Therefore the greatest improvement in symbiotic nitrogen fixation is likely to arise from simultaneous selection of both symbionts .

No abstract

This article is in response to that of Vessey (1994) who argues that the traditional, closed acetylene reduction assay can still be a valuable tool for measuring relative differences in nitrogenase activity of legumes. To counter this assertion we consider the practical uses of the traditional assay procedure in relation to real research situations. This requires the use of the assay to be considered separately in the different circumstances of pot-grown and field-grown plants. We conclude that for pot-grown legumes there are a few practical applications where the use of the traditional, closed assay procedure is valid and we accept that these can be extended by the careful use of calibrations against open, flow-through systems. However, we doubt that there are many situations where such a calibration approach would have practical advantages over using the flow-through system to obtain the actual measurements. We cannot recommend any form of the uncalibrated acetylene reduction assay for field-based studies and suggest that researchers consider the merits of simple, alternative measurements such as dry weight, yield and total nitrogen.