Abstract
The familial placement of Wightia has been controversial in the Lamiales, and the genus is currently placed in Paulowniaceae in APG IV. Phylogenetic analyses of Wightia and its close relatives in Lamiales are conducted using sequences of the complete chloroplast genomes as well as sequence data from nine chloroplast DNA regions (atpB, matK, ndhF, psbBTNH, rbcL, rps4, rps16 intron, trnL-F, and trnV-atpE) and one mitochondrial gene rps3. The maximum likelihood and Bayesian analyses do not support a close relationship between Wightia and Paulownia of Paulowniaceae; instead the enigmatic Wightia is sister to Phrymaceae with strong support in all analyses. Hence Wightia should not be placed in Paulowniaceae. Because morphological data show Wightia's affinity to both Phrymaceae and Paulowniaceae and prior nrITS data suggest its sister relationship to Paulownia of Paulowniaceae, it is likely that Wightia may have had a hybrid origin between early lineages of Phrymaceae and Paulowniaceae. It is therefore the best to exclude Wightia from Paulowniaceae and place the genus as unassigned until further nuclear data to test the hybrid hypothesis. The seven species of Paulownia constitute a monophyletic group, and Paulowniaceae is supported to be a monogeneric family, consistent with a series of morphological and floral development characters. The genus Brandisia, which was sometimes regarded as a close relative of Wightia, is supported to be nested within Orobanchaceae, as sister to Pterygiella. This sister relationship can be corroborated by fruit, seed and pollen morphological characters.