Unraveling Reticulation in the Polypodium vulgare complex

 

"The prospect before one attempting to bring anything like order out of the substantial aggregrate known as Polypodium vulgare is far from encouraging."                                                                                                  William R. Maxon, 1900                                

 

 

The Polypodium vulgare complex (Polypodiaceae) comprises a well-studied group of fern taxa whose members are cryptically differentiated morphologically and comprise a confusing and highly reticulate species cluster. Once considered a single species spanning much of northern Eurasia and North America, P. vulgare has been segregated into approximately 17 diploid and polyploid taxa as a result of cytotaxonomic work, hybridization experiments, and isozyme studies conducted during the 20th century. Despite excellent work by generations of researchers, numerous key systematic and phylogenetic questions about the complex remain. This research aims to address three areas: (1) resolving the diploid backbone of the P. vulgare complex and assessing the placement of the complex within Polypodium s.s.; (2) identifying stable morphological characters to delineate allopolyploid taxa from their progenitor species; and (3) identifying and delineating multiple origins of allopolyploid taxa. Below are examples of research in each of these three areas. 

 

Phyloreticulogram depicting hypothesized relationships within the Polypodium vulgare reticulate complex circa 2008. The bolded cladogram represents relationships among the diploid species as determined by a consensus of studies utilizing morphological, isozyme, chloroplast restriction site, and plastid sequence data. Dashed lines on the cladogram represent alternate hypotheses as suggested by one or more studies. Bolded letters A, G, and C represent the three major recognized clades of diploid species: the P. appalachianum clade, the P. glycyrrhiza clade, and the P. cambricum clade, respectively. Arrows connect polyploids to their progenitors, with dashed arrows indicating uncertain parentage. Shapes indicate each taxon’s ploidy (see inset legend). Black text, lines, and shapes indicate sexual species, whereas gray text and shapes indicate sterile hybrids. Image modified from Sigel et al. 2014, Systematic Botany.  

Phyloreticulogram depicting hypothesized relationships within the Polypodium vulgare reticulate complex circa 2008. The bolded cladogram represents relationships among the diploid species as determined by a consensus of studies utilizing morphological, isozyme, chloroplast restriction site, and plastid sequence data. Dashed lines on the cladogram represent alternate hypotheses as suggested by one or more studies. Bolded letters A, G, and C represent the three major recognized clades of diploid species: the P. appalachianum clade, the P. glycyrrhiza clade, and the P. cambricum clade, respectively. Arrows connect polyploids to their progenitors, with dashed arrows indicating uncertain parentage. Shapes indicate each taxon’s ploidy (see inset legend). Black text, lines, and shapes indicate sexual species, whereas gray text and shapes indicate sterile hybrids. Image modified from Sigel et al. 2014, Systematic Botany.  


Resolving the diploid backbone of the Polypodium vulgare complex

Previous studies of the P. vulgare complex had identified three major clades of diploid taxa (see the above phyloreticulogram), but the relationships within and among these clades remained largely unresolved. Furthermore, the relationship of the primarily-northern temperate P. vulgare complex relative to the neotropical members of Polypodium s.s. remained uncertain, with some studies questioning the monophyly of the P. vulgare complex. To address these fundamental questions, we employed phylogenetic and divergence-dating analyses of plastid sequence data from all the diploid species of the P. vulgare complex. Major finding from this work include: (1) strong support for the monophyly of the P. vulgare complex and its relationship as sister to the Neotropical P. plesiosorum group; (2) the resolution of four major diploid lineages within the P. vulgare complex; and (3) support for a late Miocene-Pliocene origin for the four major diploid lineages of the complex, with the majority of extant diploid species diversifying in the late Miocene through the Pleistocene. We were able to use these finding to reassess previous hypotheses, and to propose new hypotheses, about the historical events that shaped the diversity and current geographic distribution of the diploid species of the P. vulgare complex . This working is published in Sigel et al. 2014, Systematic Botany 

The best ML phylogram for the ten diploid taxa of the Polypodium vulgare complex, five taxa belonging to the P. plesiosorum group, and eight outgroup taxa (Table 1). ML bootstrap support values and Bayesian inference posterior probabilities are given above nodes and indicated with thickened branches (see inset legend). The bolded V and P identify the monophyletic P. vulgare complex and monophyletic P. plesiosorum group, respectively. Bolded letters A, G, C, and S indicate the major subclades of diploid species within the P. vulgare complex: the P. amorphum clade, the P. glycyrrhiza clade, the P. cambricum clade, and the P. scouleri clade, respectively. Scale bars next to each silhouette represent 2.54 cm. Image from Sigel et al. 2014, Systematic Botany. 

The best ML phylogram for the ten diploid taxa of the Polypodium vulgare complex, five taxa belonging to the P. plesiosorum group, and eight outgroup taxa (Table 1). ML bootstrap support values and Bayesian inference posterior probabilities are given above nodes and indicated with thickened branches (see inset legend). The bolded V and P identify the monophyletic P. vulgare complex and monophyletic P. plesiosorum group, respectively. Bolded letters A, G, C, and S indicate the major subclades of diploid species within the P. vulgare complex: the P. amorphum clade, the P. glycyrrhiza clade, the P. cambricum clade, and the P. scouleri clade, respectively. Scale bars next to each silhouette represent 2.54 cm. Image from Sigel et al. 2014, Systematic Botany. 


Identifying stable morphological characters

Since even before Carl Linneaus, who named Polypodium vulgare L. in Species Plantarum, taxonomists were besieged by the overwhelming intergradation of morphological variants within Polypodium s.s. This confusion reflects a dearth of stable morphological characters for the group and the morphological intermediacy of allopolyploid taxa and sterile hybrids relative to their progenitor species. For example, the allotetraploid P. calirhiza is readily mistaken for its two diploid progenitor species, P. californicum and P. glycyrrhiza, and this has led to conflicting reports about whether P. calirhiza co-occurs with P. californicum in Mexico. To address this problem, we demonstrated that a combination of spore size and rachis scale characteristics can be used to distinguish P. calirhiza from P. californicum, and that the two species have widely allopatric distributions in Mexico. Polypodium californicum is restricted to coastal regions of the Baja California peninsula and neighboring Pacific islands, whereas P. calirhiza grows at high elevations in central and southern Mexico. Interestingly, the occurrence of P. calirhiza in Oaxaca, Mexico, marks the southernmost extent of the P. vulgare complex in the Western Hemisphere (see Sigel et al. 2014, Brittonia).

Images of abaxial rachis scales for specimens of P. calirhiza (a–f) and P. californicum (g–l). Scale bars represent 100 microns. Image from Sigel et al. 2014, Brittonia.

Images of abaxial rachis scales for specimens of P. calirhiza (a–f) and P. californicum (g–l). Scale bars represent 100 microns. Image from Sigel et al. 2014, Brittonia.


Identifying and delineating multiple origins of allopolyploid taxa

Multiple hybridization events between the same diploid species can yield multiple, independently-formed lineages of an allopolyploid taxon. When these hybridization events occur in reciprocal directions, the resulting allopolyploid lineages have different maternally-inherited plastid genomes. While likely common, there are few well-documented examples of such reciprocally- derived lineages. Using a combination of maternally-inherited plastid and biparentally-inherited nuclear sequence data, we investigated the distributions and relative ages of the reciprocally-formed lineages of P. hesperium, a western North American allotetraploid fern derived from P. amorphum and P. glycyrrhiza. The reciprocally-formed lineages of P. hesperium have a striking correlation with geography. Northern populations of P. hesperium (those from Washington, Oregon, Idaho, and Montana) have P. amorphum as their maternal progenitor, whereas southern populations of P. hesperium (Utah, Colorado, New Mexico, Arizona, and Baja California) have P. glycyrrhiza as their maternal progentior. Polypodium hesperium is a ideal study system for investigating how multiple origins contribute to the genetic diversity of an allopolyploid species and the fidelity of genome evolution following allopolyploidization (see Sigel et al. 2014, American Journal of Botany). 

  Summary of the maternal identity and geographic distributions of the two reciprocally-formed lineages of Polypodium hesperium. A. Best unrooted phylogram recovered by maximum likelihood (ML) analysis of the maternally inherited plastid trnG-R data set. Red and blue circles highlight individuals of P. hesperium with maternally-inherited plastid sequences data inherited from P. amorphum and P. glycyrrhiza, respectively. Thicker branches display Bayesian inference posterior probability (BIPP) and maximum likelihood bootstrap (MLBS) support values (see inset legend). B. Geographic distribution of P. hesperium and specific collection localities of P. hesperium specimens. Gray patches represent the known geographic range of P. hesperium. Squares indicate the collection localities of particular specimens. Square color corresponds to the maternal plastid donor: red indicates P. hesperium specimens with plastids inherited from P. amorphum; blue indicates P. hesperium specimens with plastids inherited from P. glycyrrhiza. Image modified from Sigel et al. 2014, American Journal of Botany. 

 

Summary of the maternal identity and geographic distributions of the two reciprocally-formed lineages of Polypodium hesperium. A. Best unrooted phylogram recovered by maximum likelihood (ML) analysis of the maternally inherited plastid trnG-R data set. Red and blue circles highlight individuals of P. hesperium with maternally-inherited plastid sequences data inherited from P. amorphum and P. glycyrrhiza, respectively. Thicker branches display Bayesian inference posterior probability (BIPP) and maximum likelihood bootstrap (MLBS) support values (see inset legend). B. Geographic distribution of P. hesperium and specific collection localities of P. hesperium specimens. Gray patches represent the known geographic range of P. hesperium. Squares indicate the collection localities of particular specimens. Square color corresponds to the maternal plastid donor: red indicates P. hesperium specimens with plastids inherited from P. amorphum; blue indicates P. hesperium specimens with plastids inherited from P. glycyrrhiza. Image modified from Sigel et al. 2014, American Journal of Botany.