Exploring the phylogeny of the marattialean ferns

Abstract

The eusporangiate marattialean ferns represent an ancient radiation with a rich fossil record but limited modern diversity in the tropics. The long evolutionary history without close extant relatives has confounded studies of the phylogenetic origin, rooting and timing of marattialean ferns. Here we present new complete plastid genomes of six marattialean species and compiled a plastid genome dataset representing all of the currently accepted marattialean genera. We further supplemented this dataset by compiling a large dataset of mitochondrial genes and a phenotypic data matrix covering both extant and extinct representatives of the lineage. Our phylogenomic and total-evidence analyses corroborated the postulated position of marattialean ferns as the sister to leptosporangiate ferns, and the position of Danaea as the sister to the remaining extant marattialean genera. However, our results provide new evidence that Christensenia is sister to Marattia and that M. cicutifolia actually belongs to Eupodium. The apparently highly reduced rate of molecular evolution in marattialean ferns provides a challenge for dating the key phylogenetic events with molecular clock approaches. We instead applied a parsimony-based total-evidence dating approach, which suggested a Triassic age for the extant crown group. The modern distribution can best be explained as mainly resulting from vicariance following the breakup of Pangaea and Gondwana. We resolved the fossil genera Marattiopsis, Danaeopsis and Qasimia as members of the monophyletic family Marattiaceae, and the Carboniferous genera Sydneia and Radstockia as the monophyletic sister of all other marattialean ferns.

Inferring the Total-Evidence Timescale of Marattialean Fern Evolution in the Face of Model Sensitivity

“…Our analyses allow for the long stem branch to be broken up by fossil accessions, which may additionally provide important information about the polarity of morphological characters (i.e., which character states are ancestral for crown Marattiaceae), potentially allowing for more reliable inference of the "root" position for this crown clade (Doyle and Donoghue 1987;Gauthier et al 1988;Donoghue et al 1989;Huelsenbeck 1991;Smith 1998;Wills and Fortey 2000; Mongiardino Koch and Parry 2020). Nonetheless, our results further support the growing consensus (Pryer et al 2001;Qiu et al 2007;Murdock 2008b;Rai and Graham 2010;Lehtonen et al 2020) that Danaea is sister to the remaining extant lineages; it shares synapomorphic states of the extant clade (e.g., the presence of stipules), and has apomorphic states (e.g.,…”

“…However, support is relatively weak for these relationships in our results as well as in Murdock (2008b). Lehtonen et al (2020), while likewise having only weak support, resolve the third possible relationship: Christensenia sister to Marattia, and that clade sister to Angiopteris. This phylogenetic uncertainty is mirrored in previous morphological cladistic studies (see Hill and Camus 1986), as well as the samplingfocused analyses of Rothwell et al (2018b), which have supported each of these positions for Christensenia, in addition to placing it sister to Danaea or to all other extant genera.…”

“…6). Given that empirical datasets 1195 frequently exhibit imbalanced taxon sampling schemes, 1196 and some degree of "diversified" or "deep-node" sam- Rothwell et al 2018b;Lehtonen et al 2020). Among the extant taxa, the main point of uncertainty remains the position of Christensenia.…”

“…The extensive fossil record of the Marattiales, in its rich-156 ness, completeness, and temporal extent (Millay 1997;well et al 2018b; see also Lehtonen et al 2020, which was 171 published after our analyses were completed).…”

“…This result only holds when temporal data are incorporated; in the unrooted analyses (Fig. S69) some of these fossils are resolved among the extant species, as they were in, e.g., Rothwell et al (2018b) and Lehtonen et al (2020). (2017) and Smith et al (2010), respectively.…”