Aim Studying plant–soil interactions of introduced species in different parts of their global range could assist in managing biological invasions by elucidating the level of host specificity of key mutualists. We assessed the role of the soil microbial community (with an emphasis on symbiotic nitrogen‐fixing bacteria, collectively termed rhizobia) in determining cross‐continental invasion success of five woody legume species. Location Australia. Methods For each species, we compared growth of plants in soils from their native and non‐native ranges using a glasshouse study, a soil dilution method (most probable number) and T‐RFLP to assess rhizobial abundance and community composition, respectively. Results Acacia longifolia was the only species that had significantly larger above‐ground biomass when grown in soils from its non‐native range. Rhizobial abundance was equally high across species and ranges, indicating plants are unlikely to be limited by soil rhizobial abundance in non‐native ranges. Acacia cyclops, A. saligna and Paraserianthes lophantha formed associations with different rhizobial communities in non‐native vs. native range soils. Acacia longifolia and A. melanoxylon associated with similar rhizobial communities in their native and non‐native ranges, suggesting that rhizobia may have been accidentally introduced into their novel range with seeds or seedlings. Main conclusions Invasive success of these five legume species is not constrained by the abundance of rhizobia in novel ranges for established legume populations, at least within Australia. Although differences in rhizobial community composition were evident between the native and non‐native ranges for three of the five species, these were not associated with differences in plant growth. Increased above‐ground biomass of A. longifolia when grown in soil from its non‐native range suggests that invasive success of this species may be associated with differences in the non‐rhizobial components of soil microbial communities in the novel range. This information could assist in management practises by facilitating a more instructive and effective screening for invasiveness.

No abstract

Methods 42NifH genes were amplified from rhizosphere soils collected from beneath each species 43 (multiple populations) within their native and introduced range and directly from 44 nodules collected from plants previously grown in the glasshouse using field-collected 45 soil as inoculum. NifH gene sequences from soils and nodules were 454 46 pyrosequenced and assigned to taxonomic groups based on nifH consensus taxonomy. 47 48 Results 3We found no difference in the NFB community of soils or nodules between native and 50 introduced ranges across the five species, suggesting that these legumes encounter 51 similar NFB communities in soils across Australia. Bradyrhizobium was the most 52 abundant rhizobial genus present in both soils and nodules. Bradyrhizobium species 53 found in nodules were significantly different across the ranges for A. longifolia. 54 55 Main conclusions 56The results indicate that these invasive legumes have similar nitrogen fixing bacterial 57 communities in their rhizosphere and nodules across Australia, with the exception of 58A. longifolia. This species has diverse Bradyrhizobium genotypes in its nodules 59suggesting that A. longifolia may be a more generalist host compared to the other four 60 legumes. Thus it is unlikely that the invasive success of these legumes is constrained 61 Soil microbes are increasingly acknowledged to play significant roles in invasion 68 outcomes for many plant species (Inderjit & Cahill, 2015;Vestergård et al., 2015). (Wandrag et al., 2013). 92Broad symbiotic promiscuity and ability to nodulate at low rhizobial abundance have 93 been described as significant advantages for invading legumes (Parker, 2001; Perez-94 Fernanndez & Lamont, 2003;Rodríguez-Echeverria et al., 2011). Previous studies 95 have shown that some invasive woody legumes are able to readily nodulate (Lafay & 96 Burdon, 2006) and associate with novel bacterial communities in their exotic ranges 97 (Marsudi et al., 1999; Amrani et al., 2010; Callaway et al., 2011; Ndlovu et al., 98 2013). 99Despite the evidence for promiscuity for some host species, there are also reports 100showing that some invasive legumes in their invasive range have specificity towards 101 rhizobia from their native range (Chen et al., 2005). Such results were reported for A. 102 longifolia in Portugal which was found to associate with rhizobial communities that 103 were very similar to those from A. longifolia's native range in south-east Australia 104 (Rodríguez-Echeverria, 2010). Thus, there appears to be considerable variation 105 between legume hosts and their symbiotic associations across introduced and native 106 ranges. 107There is evidence to suggest that invasive acacias (sensu Richardson et al. 2011) 108 could be more promiscuous, i.e., they are able to associate with more diverse rhizobia, 109 than non-invasive acacias (Klock et al., 2015). However, there is also evidence to 110 suggest that some acacias (e.g., A. cyclops, A. pycnantha) are able to non-specifically 111 nodulate with both fast-and...

No abstract

10Frequency and intensity of disturbance is projected to increase for many ecosystems globally, 11 with uncertain consequences, particularly when disturbances occur in rapid succession. We 12 quantified community response (fifty-two shrub species and the tree Eucalyptus todtiana) to a 13 severe hailstorm followed two months later by prescribed fire for a Mediterranean-type 14 shrubland in southwestern Australia. Partial overlap of hailstorm path and fire perimeter 15 provided a unique opportunity to compare storm and fire effects along a storm severity gradient 16(high-moderate-none) with and without fire. We quantified disturbance severity (bark and 17 canopy removal, scorch height) and subsequent response (resprouting type, quantity and quality,