Variation in the effectiveness of symbiotic associations between native rhizobia and temperate Australian legumes: interactions within and between genera

1. Tests of the effectiveness of rhizobia isolated from several common Acac ia species showed significant variation among host populations in plant gro wth but no indication that plant performance depended on which Acacia speci es the rhizobial isolate was derived from. Variation in effectiveness was p rimarily between rhizobial isolates within a host species rather than acros s rhizobial isolates from different hosts. No interactions were found betwe en host population, host origin of rhizobial strain and rhizobial isolate. 2. A study using a range of less common Acacia species showed significant v ariation in performance of several species (A. silvestris, A. nano-dealbata , A. trachyphloia, A. filicifolia, A. leucoclada) as a function of the inoc ulating rhizobial strain; two other species (A. cangaiensis, A. glaucocarpa ) were marginally non-significant. For A. filicifolia and A. silvestris, gr owth was significantly better when plants were inoculated with their own is olates vs. those from other host species. 3. Comparisons of the performance of a third set of Acacia species found th at several common species (A. melanoxylon, A. mearnsii, A. irrorata, A. bin ervata) varied little in their response to particular rhizobial isolates; f or these species, most host-rhizobial combinations resulted in similar grow th responses. Several less common species showed significant variation in p erformance across the rhizobial isolates tested. 4. Preliminary investigations of host specificity among legume genera showe d significant variation with respect to host performance as a function of t he rhizobial isolate with which plants were inoculated. Some species (e.g. A. mearnsii and Indigofera australis) were variable in their responses, whi le others (e.g. Hardenbergia violacea and A. melanoxylon) performed well wi th most isolates. 5. A comparison of the effectiveness of several Acacia-derived strains on t heir own hosts vs. their effectiveness on the other legume genera in the st udy showed that, in general, the Acacia strains performed well on other gen era. However, on some host species (e.g. I. australis) the Acacia strains w ere less effective than the Indigofera strains, suggesting some host specif icity. 6. While there was a general correlation at the among-species level in Acac ia-rhizobia interactions (i.e. isolates that performed well on one host spe cies often performed well on others), growth performance with any particula r host-rhizobial combination was more difficult to predict than within-spec ies comparisons. Our data also suggest that, at least for the genus Acacia, rare species are likely to exhibit a greater degree of host specificity th an hosts that are widely distributed, and re-establishment may require spec ial attention with respect to rhizobial strains that are used. 7. Overall, these results have clear implications for applied projects focu sed on the restoration of degraded habitats or abandoned farmland. Simply r eplanting native species into such areas may be unsuccessful unless appropr iate rhizobial strains are also reintroduced. Our results suggest that when no strains for a particular host species are available, strains from its c losest relative will have the highest probability of success.