Why the Monophyly of Nymphaeaceae Currently Remains Indeterminate: An Assessment Based on Gene-Wise Plastid Phylogenomics

Abstract

The monophyly of Nymphaeaceae (water lilies) represents a critical question in understanding the evolutionary history of early-diverging angiosperms. A recent plastid phylogenomic investigation claimed new evidence for the monophyly of Nymphaeaceae, but its results could not be verified from the available data. Moreover, preliminary gene-wise analyses of the same dataset provided partial support for the paraphyly of the family. The present investigation aims to re-assess the previous conclusion of the monophyly of Nymphaeaceae under the same dataset and to determine the congruence of the phylogenetic signal across different plastome genes and data partition strategies. To that end, phylogenetic tree inference is conducted on each of 78 protein-coding plastome genes, both individually and upon concatenation, and under four data partitioning schemes. Moreover, the possible effects of various sequence variability and homoplasy metrics on the inference of specific phylogenetic relationships are tested using multiple logistic regression. Differences in the variability of polymorphic sites across codon positions are assessed using parametric and non-parametric analysis of variance. The results of the phylogenetic reconstructions indicate considerable incongruence among the different gene trees as well as the data partitioning schemes. The results of the multiple logistic regression tests indicate that the fraction of polymorphic sites of codon position 3 has a significant effect on the recovery of the monophyly of Nymphaeaceae. Taken together, these results indicate that the monophyly of Nymphaeaceae currently remains indeterminate, and that specific phylogenetic conclusions are strongly dependent on the precise plastome gene, data partitioning scheme, and codon position evaluated. In closing, I discuss the importance of archiving all data of an investigation in publicly accessible data repositories, along with sufficient details to replicate the published results, and provide recommendations on future plastid phylogenomic investigations of Nymphaeales.

Generalist Pollen-Feeding Beetles during the Mid-Cretaceous

“…The Nymphaeaceae (water lily) together with the Cabombaceae (water shields) and the Hydatellaceae, constitutes the Nymphaeales, a clade of aquatic plants that is a sister group or adjacent paraphyletic group to extant Amborellaceae (Friis et al, 2001;Soltis et al, 2018), the phylogenetically basalmost angiosperm lineage (Friis et al, 2001;APG, 2009). However, at present its monophyly represents a critical question (Gruenstaeudl, 2019). Compared with other angiosperm plant lineages of the Cretaceous, the water lilies have a modest fossil record.…”

Historical biogeography of Vochysiaceae reveals an unexpected perspective of plant evolution in the Neotropics

“…Analyses using the MSC method were performed to explore the phylogenetic signal variation and to detect potential controversial nodes. Recent studies have shown the importance of exploring incongruence between gene trees and species trees when dealing with data from the plastome (Ruhlman and Jansen, 2018;Gonçalves et al, 2019;Gruenstaeudl, 2019;Walker et al 2019). Gene trees for 78 plastid protein-coding genes used as input for ASTRAL-III were inferred using IQ-TREE version 1.5.5 (Nguyen et al, 2015).…”

Plastid phylogenomics of the Gynoxoid group (Senecioneae, Asteraceae) highlights the importance of motif-based sequence alignment amid low genetic distances

“…The authors reported that the concatenation of all plastid coding regions produced highly supported phylogenies that were nonetheless incongruent to individual plastid gene trees. Similarly, Gruenstaeudl (2019) detected phylogenetic incongruence across different loci of the plastid genome in a phylogenomic investigation of water lilies and relatives. Walker et al (2019) reported gene tree conflict among various plastid genes based on a broad sampling of angiosperm plastid genomes and noted numerous strongly supported but conflicting nodes between different gene trees.…”

“…Hence, many researchers assume that the plastid genome evolves as a single linkage unit (Bock, 2007), with different genomic regions sharing the same evolutionary history (e.g., Lu et al, 2018). However, recent studies have reported widespread phylogenetic incongruence between different regions of the plastid genome (e.g., Goncalves et al, 2019;Gruenstaeudl, 2019;Walker et al, 2019), which indicates that the plastid genome may not represent a homogeneous genetic locus. This phylogenetic incongruence may partially be the result of the different mutation rates and selective constraints across the plastid genome, which is illustrated by the co-existence of the relatively slowly evolving gene rbcL, the more rapidly evolving gene matK with nearly equal site rates in all three codon positions, and the even faster evolving non-coding markers trnL intron, trnT-L intergenic spacer, and trnL-F intergenic spacer in the same genome (Müller et al, 2006).…”

“…The different regions of the plastid genome exhibit different molecular constraints and, thus, a different utility for phylogenetic reconstructions. The coding regions of the plastid genome have traditionally been used for the reconstruction of deep-level phylogenetic relationships Xi et al, 2012;Walker et al, 2019), but these may be biased when genes with incongruent phylogenetic signal coexist in the genome (e.g., Gruenstaeudl, 2019;Goncalves et al, 2019;Walker et al, 2019). Goncalves et al (2020), suggested to infer phylogenetic relationships on individual genes and to compare the resulting topologies to the reconstructions based on the concatenation of all coding regions.…”