Data from: An unusual new rheophytic species of the liverwort genus Cephaloziella from Panama
Data files
Jan 08, 2026 version files 163.62 KB
-
Cephaloziellaceae_ITS_alignment_may25.nex
76.65 KB
-
Cephaloziellaceae_ITS_mod.tree
4.05 KB
-
Cephaloziellaceae_rbcL_alignment_may25.nex
76.25 KB
-
Cephaloziellaceae_rbcL_mod.tree
3.32 KB
-
README.md
3.34 KB
Abstract
During bryophyte surveys in the Antón River watershed, Coclé Province, Panama, we discovered a minute rheophytic liverwort representing an undescribed species of Cephaloziellaceae. Phylogenetic analyses based on nuclear ITS1–2 and plastid rbcL DNA markers placed the new species within Cephaloziella, where it resolved in a clade with C. microphylla, C. kiaeri, and C. recurvifolia in the ITS tree, and with C. microphylla alone in the rbcL tree. We describe the species as Cephaloziella minutiloba A.M. Sierra, Gradst. & Gudiño sp. nov. The new species is morphologically distinct by its minute size, flagelliform shoots, scattered violet rhizoids, succubous and highly unequally conduplicate leaves with a prominent ventral lobe and a strongly reduced dorsal lobe bent over the ventral lobe, and crenate-dentate leaf margins. The gynoecia on short ventral branches are made up of 2–3 series of highly connate, crenate bracts and a 5-plicate perianth with a fringed mouth. This new species expands our understanding of morphological diversity and rheophytic adaptation within Cephaloziella. The morphological affinities of the species are discussed, and the importance of integrative taxonomic approaches in liverwort systematics is emphasized.
Dataset DOI: 10.5061/dryad.b2rbnzsv0
Citation
Adriel M. Sierra, S. Robbert Gradstein, and José Gudiño L "An Unusual New Rheophytic Species of the Liverwort Genus Cephaloziella from Panama," Systematic Botany 50(3), 185-194, (18 December 2025). https://doi.org/10.1600/036364425X17637402353547
Description of the data and file structure
Associated dataset for: An unusual new rheophytic species of the liverwort genus Cephaloziella from Panama
Adriel M. Sierra1, S. Robbert Gradstein2 & José Gudiño L.3
1 Département de Biologie, Université Laval, Québec, (QC), G1V 0A6, Canada.
2 Meise Botanic Garden, 1860 Meise, Belgium
3 Smithsonian Tropical Research Institute, Summit Herbarium (SCZ), PO Box 0843-03092, Ancón, Panama.
Files and variables
File: Cephaloziellaceae_ITS_alignment_may25.nex
Description: Alignment of 58 sequences of the nuclear gene ITS comprises 1251 base pairs. Nuclear ribosomal internal transcribed spacer (ITS1–2) region nrITS1–5.8S–ITS2, sequenced with the primers Bryo18SF and Bryo26SR.
File: Cephaloziellaceae_rbcL_alignment_may25.nex
Description: Alignment of 48 sequences of the chloroplast region rbcL comprises 1514 base pairs. The rbcL sequence where generated with the primers rbcL-210-F and rbcL-1200-R.
File: Cephaloziellaceae_ITS_mod.tree
Description: ITS1–2 maximum likelihood consensus tree.
File: Cephaloziellaceae_rbcL_mod.tree
Description: rbcL maximum likelihood consensus tree.
Code/software
Sequences were edited and aligned with Geneious Prime version 2024.0.4 (https://www.geneious.com). Alignments were performed automatically using the default settings (65 % similarity, gap open penalty = 12, gap extension penalty = 3, refinement iterations = 2) and were later checked manually.
Phylogenetic analyses—Gene trees were inferred in IQ-tree using 1000 ultrafast bootstrap replicates, a maximum likelihood inference method with specific evolutionary models for each gene (ITS: GTR+F+G4; and rbcL: TIM+F+G4) using ModelFinder (Kalyaanamoorthy et al. 2017; Minh et al. 2013). The phylogenetic tree with ML result was manipulated and visualized using the packages ape and phylotools (Zhang 2017; Paradis and Schliep 2019) in the R software (R Core team 2024). Bootstrap values ≥ 70 were regarded as moderate and ≥ 85 as good support for the new species (Erixon et al. 2003). Bayesian inference was performed on the software MrBayes v.3.2.6 (Ronquist and Huelsenbeck 2003, Ronquist et al. 2012). Two parallel Markov Chain Monte Carlo (MCMC) runs, adding up to ten million generations, were conducted, with each run containing four chains with default priors on most parameters. Trees and estimated parameter values were sampled every 1,000 generations, thus obtaining a total of 10,000 samples from which the first 1,000 (10%) were discarded as burn-in. Convergence of runs was verified using Tracer 1.6 (Rambaut et al. 2014). A majority-rule consensus tree was computed to calculate the Bayesian posterior probability (BPP), in which values of BPP ≥ 0.95 were considered significant.