Limits...
The structure of legume – rhizobium interaction networks and their response to tree invasions

View Article: PubMed Central - PubMed

ABSTRACT

We provide data on how legume-rhizobia interaction webs react to invasions by exotic legumes. This is the first study of its kind and found that general hypotheses derived from above-ground mutualistic webs may not hold for below-ground counterparts. Specifically, we found that legume-rhizobia interactions at the community level are highly specialised resulting in strongly modular webs, which are not nested, and that invasive legumes do not infiltrate existing native webs but rather form unique and novel modules in webs.

No MeSH data available.


Related in: MedlinePlus

Networks representing legume–rhizobium interactions across a gradient of acacia invasion (uninvaded, semi invaded and heavily invaded sites) for bacterial taxa defined as individual genotypes (see Figs. 1 and 2) and at the 98% 16S rDNA sequence similarity level. Rows represent plant taxa (red species names: invasives, blue species names: natives). Columns represent nodule associated bacterial taxa (B – beta rhizobia in the genus Burkholderia, Ar, alpha rhizobia in the Rhizobium clade (Fig. 2, clade B); Ab, alpha rhizobia in the Bradyrhizobium clade (Fig. 2, clade A)). Increasing intensity of blue represents the increasing frequency of species interactions. Red boxes represent modules identified by the weighted modularity approach of Dormann and Strauss (2014). The invasive and native plant species always occupy distinct modules.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4940501&req=5

plw038-F4: Networks representing legume–rhizobium interactions across a gradient of acacia invasion (uninvaded, semi invaded and heavily invaded sites) for bacterial taxa defined as individual genotypes (see Figs. 1 and 2) and at the 98% 16S rDNA sequence similarity level. Rows represent plant taxa (red species names: invasives, blue species names: natives). Columns represent nodule associated bacterial taxa (B – beta rhizobia in the genus Burkholderia, Ar, alpha rhizobia in the Rhizobium clade (Fig. 2, clade B); Ab, alpha rhizobia in the Bradyrhizobium clade (Fig. 2, clade A)). Increasing intensity of blue represents the increasing frequency of species interactions. Red boxes represent modules identified by the weighted modularity approach of Dormann and Strauss (2014). The invasive and native plant species always occupy distinct modules.

Mentions: Regardless of how bacterial taxa were delineated (i.e. from genotype to 95% DNA sequence similarity levels) interaction networks at all sites were always significantly modular (results reported for genotype and 98% levels in Table 2 and Figure 4, [see Supporting Information] with native and invasive legumes always occupying different modules (Fig. 4, [see Supporting Information]). At the bacterial genotype level each of the five legume species at each site formed a module with a unique set of bacteria (Fig. 4), while at the 98% DNA similarity level four modules were recognized at the uninvaded site and three at each of the invaded sites (two modules containing invasive legumes and one containing the natives). At the 95% DNA similarity level networks comprise two modules, one containing the natives and the other the invasives [see Supporting Information]. Network topology did not differ markedly across the invasion gradient (Table 2, [see Supporting Information]) and species level metrics did not differ significantly between sites [see Supporting Information]. All networks were not significantly nested, weakly connected (connectance < 0.55) with an even spread of interactions (IE > 0.57) and high levels of network specialization (H’2 > 0.61). Species level metrics did not differ between native and invasive legumes (degree: natives (n) = 4.9 ± 1.5, invasives (i) = 3.8 ± 1.0, F1,6 = 1.25 ns; effective partners: n = 4.1 ± 1.1, i = 2.7 ± 1.2, F1,6 = 1.25 ns; specialization: n = 0.85 ± 0.10, i = 0.95 ± 0.03, F1,6 = 2.49 ns).Figure 4.


The structure of legume – rhizobium interaction networks and their response to tree invasions
Networks representing legume–rhizobium interactions across a gradient of acacia invasion (uninvaded, semi invaded and heavily invaded sites) for bacterial taxa defined as individual genotypes (see Figs. 1 and 2) and at the 98% 16S rDNA sequence similarity level. Rows represent plant taxa (red species names: invasives, blue species names: natives). Columns represent nodule associated bacterial taxa (B – beta rhizobia in the genus Burkholderia, Ar, alpha rhizobia in the Rhizobium clade (Fig. 2, clade B); Ab, alpha rhizobia in the Bradyrhizobium clade (Fig. 2, clade A)). Increasing intensity of blue represents the increasing frequency of species interactions. Red boxes represent modules identified by the weighted modularity approach of Dormann and Strauss (2014). The invasive and native plant species always occupy distinct modules.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4940501&req=5

plw038-F4: Networks representing legume–rhizobium interactions across a gradient of acacia invasion (uninvaded, semi invaded and heavily invaded sites) for bacterial taxa defined as individual genotypes (see Figs. 1 and 2) and at the 98% 16S rDNA sequence similarity level. Rows represent plant taxa (red species names: invasives, blue species names: natives). Columns represent nodule associated bacterial taxa (B – beta rhizobia in the genus Burkholderia, Ar, alpha rhizobia in the Rhizobium clade (Fig. 2, clade B); Ab, alpha rhizobia in the Bradyrhizobium clade (Fig. 2, clade A)). Increasing intensity of blue represents the increasing frequency of species interactions. Red boxes represent modules identified by the weighted modularity approach of Dormann and Strauss (2014). The invasive and native plant species always occupy distinct modules.
Mentions: Regardless of how bacterial taxa were delineated (i.e. from genotype to 95% DNA sequence similarity levels) interaction networks at all sites were always significantly modular (results reported for genotype and 98% levels in Table 2 and Figure 4, [see Supporting Information] with native and invasive legumes always occupying different modules (Fig. 4, [see Supporting Information]). At the bacterial genotype level each of the five legume species at each site formed a module with a unique set of bacteria (Fig. 4), while at the 98% DNA similarity level four modules were recognized at the uninvaded site and three at each of the invaded sites (two modules containing invasive legumes and one containing the natives). At the 95% DNA similarity level networks comprise two modules, one containing the natives and the other the invasives [see Supporting Information]. Network topology did not differ markedly across the invasion gradient (Table 2, [see Supporting Information]) and species level metrics did not differ significantly between sites [see Supporting Information]. All networks were not significantly nested, weakly connected (connectance < 0.55) with an even spread of interactions (IE > 0.57) and high levels of network specialization (H’2 > 0.61). Species level metrics did not differ between native and invasive legumes (degree: natives (n) = 4.9 ± 1.5, invasives (i) = 3.8 ± 1.0, F1,6 = 1.25 ns; effective partners: n = 4.1 ± 1.1, i = 2.7 ± 1.2, F1,6 = 1.25 ns; specialization: n = 0.85 ± 0.10, i = 0.95 ± 0.03, F1,6 = 2.49 ns).Figure 4.

View Article: PubMed Central - PubMed

ABSTRACT

We provide data on how legume-rhizobia interaction webs react to invasions by exotic legumes. This is the first study of its kind and found that general hypotheses derived from above-ground mutualistic webs may not hold for below-ground counterparts. Specifically, we found that legume-rhizobia interactions at the community level are highly specialised resulting in strongly modular webs, which are not nested, and that invasive legumes do not infiltrate existing native webs but rather form unique and novel modules in webs.

No MeSH data available.


Related in: MedlinePlus