The infectivity of the soybean symbiont Rhizobiumjaponicum changed two-to fivefold with culture age for strains 110 ARS, 138 Str Spc, and 123 Spc, whereas culture age had relatively little effect on the infectivity of strains 83 Str and 61A76 Str. Infectivity was measured by determining the number of nodules which developed on soybean primary roots in the zone which contained developing and preemergent root hairs at the time of inoculation. Root cells in this region of the host root are susceptible to Rhizobium infection, but this susceptibility is lost during acropetal development and maturation of the root cells within a period of 4 to 6 h (T. V. Bhuvaneswari, B. G. Turgeon, and W. D. Bauer, Plant Physiol. 66:1027-1031, 1980. Profiles of nodulation frequency at different locations on the root were not affected by the age of the R. japonicum cultures, indicating that culture age affected the efficiency of Rhizobium infection rather than how soon infections were initiated after inoculation. Inoculum dose-response experiments also indicated that culture age affected the efficiency of infection. Two strains, 61A76 Str and 83 Str, were relatively inefficient at all culture ages, particularly at low inoculum doses. Changes in infectivity with culture age were reasonably well correlated with changes in the proportion of cells in a culture capable of binding soybean lectin. Suspensions of R. japonicum in water were found to retain their viability and infectivity. nm per 100 ml), and grown to the desired growth phase for use as inocula. Inocula were prepared by straight dilution of cultures with sterile water. Rhizobium cell numbers were determined by direct counting in Pe-443 on August 5, 2020 by guest

Three Rhizobiumjaponicum strains and two slow-growing cowpea-type Rhizobium strains were found to remain viable and able to rapidly nodulate their respective hosts after being stored in purified water at ambient temperatures for periods of 1 year and longer. Three fast-growing Rhizobium species did not remain viable under the same water storage conditions. After dilution of slow-growing Rhizobium strains with water to 103 to 105 cells ml-', the bacteria multiplied until the viable cell count reached levels of between 106 and 107 cells ml-1. The viable cell count subsequently remained fairly constant. When the rhizobia were diluted to 107 cells ml-1, they did not multiply, but full viability was maintained. If the rhizobia were washed and suspended at 109 cells ml-', viability slowly declined to 107 cells ml-' during 9 months of storage. Scanning electron microscopy showed that no major morphological changes took place during storage. Preservation of slow-growing rhizobia in water suspensions could provide a simple and inexpensive alternative to current methods for the preservation of rhizobia for legume inoculation.

SUMMARYThe adhesion of rhizobia to surfaces of clover roots was examined by an indirect plate-counting assay and phase-contrast microscopy. The number oiRhizobium trifolii cells attached to clover root segments increased in approximately linear fashion during the first hour of incubation, but did not change appreciably thereafter. The addition of 30 mM-2-deoxy-D-glucose, which effectively inhibits binding of clover root lectin, did not promote the release of previously attached bacteria nor inhibit subsequent attachment to either root segments or root hairs. Rhizobia of several heterologous species attached to clover roots in numbers comparable to those of strains of R. trifolii, the homologous species. These results indicate that rhizobia have effective mechanisms of adhesion to non-host roots and that clover lectin contributed little or nothing to attachment under the conditions examined. IN TR O D U CTIO NThere has been considerable interest in the identification of the molecular and cellular factors that determine how rhizobia and their host plants interact to establish biologically specific symbiotic associations. About a decade ago, Bohlool & Schmidt (1974) provided correlative evidence that the binding of host plant lectins to carbo hydrate receptors from Rhizobium cells might play an important role in recognition interactions between the symbionts. Dazzo & Hubbell (1975) subsequently found evidence for the presence of a lectin in clover that bound specifically to infective isolates of R. trifolii, but did not bind to non-infective isolates or to strains of rhizobia that nodulated other legume hosts. Binding of the clover lectin, trifoliin A, to capsular receptors of the rhizobia was specifically inhibited by the addition of 2-deoxy-D-glucose (Dazzo & Hubbell, 1975). Addition of 2-deoxyglucose also released trifoliin A from roots (Dazzo, Yanke & Brill, 1978). Dazzo & Hubbell (1975) proposed that Rhizobium attachment to the host root surface was mediated by binding of the clover lectin to sur face receptors on both of the symbiotic partners, and suggested that such lectinmediated attachment was instrumental in determining recognition and host specificity. This model has had a considerable influence on subsequent enquiry in the field.Dazzo, Napoli & Hubbell (1976) later investigated both the specificity of Rhizobium attachment to clover roots and the role of clover lectin in such attachment. They reported that infective isolates of R. trifolii attached about 10-fold more readily to clover root hairs than several non-infective isolates or three strains of a heterologous species. They also reported that the addition of 2-deoxyglucose caused a specific (Stacey, Paau & Brill, 1980; Paau, Leps & Brill, 1981; Kato, Maruyama & Nakamura, 1981).A number of investigations, on the other hand, showed that rhizobia could attach in substantial numbers to roots of non-host legumes (Broughton, van Egeraat & Lie, 1980; Chen & Phillips, 1976; Peters & Alexander, 1966), as well as to roots of non-legumes (Shimshick & ...