In the selection of acid-tolerant Rh&obium meliloti, procedures for the collection and isolation of rhizobia, and the assessment of acid tolerance, have not been critically evaluated. Such procedures form the basis of this study. Root nodules were collected from Medicago spp. found growing on acid soil in Sardinia. Their encumbent bacteria were isolated directly onto media adjusted over a range ofpH values, and then assessed for acid tolerance in both the laboratory and field. Strains of Rh&obium meliloti isolated onto low pH media were, in general, more acid-tolerant than sister isolates from high pH media, when tested in both the laboratory and field. Dilution (10 or 100 fold) of the inocula used in the laboratory assessment did not greatly influence the rating derived, although there was some effect of bacterial colony type on growth rating. The link between polysaccharide production and acid tolerance was not strong. There was a poor correlation between the growth ratings derived from the laboratory screening and acid tolerance as expressed in the field.

Several strains of Rhizobium meliloti that originated from acid soils in Sardinia, Italy, were markedly superior in colonizing a moderately acid loamy sand (pH 5.0 in 1:5 0.01 M CaCl2) than two Australian commercial inoculant strains (U45 and CC169), and a group of strains that originated from alkaline soils in Syria and Iraq. Six Medicago hosts also varied greatly in their ability to achieve nodulation in this soil. M. polymorpha and M. murex were far superior in this respect to M. littoralis, M. truncatula and M. tornata. The most acid-tolerant strains of R. meliloti, WSM419 and WSM413, were able to nodulate a high proportion of plants of M. polymorpha and M. murex sown in the second year between 11 and 20 cm from the point of introduction of the rhizobia into the soil the previous year. It is suggested that these more saprophytically competent isolates of R. meliloti, combined with the species of Medicago more able to nodulate readily in acid soil, will extend the range of soils suitable for successful regenerative growth of these species.

Burrs of 3 cultivars of subterranean clover (Trifolium subterraneum) and 1 cultivar each of burr medic (Medicago polymorpha) and barrel medic (M. truncatula), which had experienced 1 summer at the soil surface, were placed on the soil surface and at depths of 2, 6 and 10 cm in the soil. The numbers of residual hard seeds were determined each year for up to 4 years. There was a marked reduction in the rate of seed softening in all 3 clover cultivars with increasing depth of burial. Whereas <20% of the seeds of the hardest seeded clover cultivar, Nungarin, survived 3 years at the soil surface, there was no significant decline in seed numbers during 4 years of burial at 10 cm. Even with cv. Geraldton, in which only 5% of seeds remained after 1 year of placement at the soil surface, 75% of seeds survived 4 years of burial at 10 cm. Hard seeds of both medic varieties were considerably more resilient than clover seeds at the soil surface, particularly during the first summer following seed set. However, burial had much less effect on their longevity, with no significant effect of burial to 2 cm in either medic, or of burial to 6 cm in the case of barrel medic. These results support earlier findings which showed that tillage operations associated with crop establishment which result in the burial of substantial proportions of subterranean clover seeds can lead to useful soil seed reserves. The much lesser effect of burial on seed softening of the medics, compared with subterranean clover, suggests that tillage operations will be less advantageous to medic persistence in leys.

No abstract

The short-term (within-year) dynamics of the softening of hard seeds in a number of accessions of Trifolium subterraneum L., T. glomeratum L., and Medicago polymorpha L. were monitored in the field. There were distinct differences in the patterns of seed softening between and within species and between years. Seed softening was accurately described by logistic curves with calculated half-lives of hard seeds (within a given year) a good indicator of differences in the softening patterns between species and accessions. T. subterraneum cv. Nungarin softened most rapidly over summer, ceasing by March (half-life in the first year 45 days), whereas M. polymorpha cv. Serena and 2 accessions of T. glomeratum softened mainly during the autumn (half-lives of 126, 104, and 136 days, respectively) First year half-lives of 4 other accessions of T. subterraneum ranged from 64 to 79 days. The results showed that large seeds were more likely to soften in the first year than were small seeds. The different patterns can be explained using Taylor's 2-stage model of seed softening. The implications of different patterns are discussed in terms of adaptation to a Mediterranean environment. T. glomeratum and M. polymorpha cv. Serena are considered to have a short-term pattern of seed softening well adapted to an environment where false breaks to the growing season are likely. The pattern of T. subterraneum is considered to be less well adapted to such an environment. However, variation within the species indicates the potential for selection of better adapted varieties. The inclusion of the short-term seed softening pattern as a selection criterion for pasture legumes is recommended.