Figure 3 displays the synthesis of 34 alphaproteobacterial genes (atp1, atp6, atp9, cob, cox2, cox3, nad1, nad2, nad3, nad4, nad4l, nad5, nad6, nad7, nad8, nad11, rpl2, rpl5, rpl6, rpl11, rpl14, rpl16, rpoA, rpoB, rpoC, rps7, rps10, rps12, rps13, rps14, rps19, sdh2, sdh3 and tufA). The proposed vertical-inheritance backbone representing the concatenation tree is shown in dark blue, with the line thickness of an internal branch corresponding to the frequency of its support across the whole dataset. Support was considered significant when clades received > 50% bootstrap support. Putative LGT events are in orange, connecting donors (circles) with recipients (arrowheads); where there are multiple possible donor candidates, these converge onto a double arrowhead. This happens when the clade founded by a past LGT donor may have subsequently had its species membership obfuscated by later exchanges of genetic material, yielding a non-reference assemblage of species labels in a presumed lineage. Where the apparent donor of a gene falls outside of the taxa included in the analysis, one is created as a basal group taxon, indicated in light blue. In order to avoid graphical congestion, branches in the tree may be artificially extended, as dotted segments.