Ostariophysi is a superorder of bony fishes including more than 10,300 species in 1,100 genera and 70 families. This superorder is traditionally divided into five major groups (orders): Gonorynchiformes (milkfishes and sandfishes), Cypriniformes (carps and minnows), Characiformes (tetras and their allies), Siluriformes (catfishes), and Gymnotiformes (electric knifefishes). Unambiguous resolution of the relationships among these lineages remains elusive, with previous molecular and morphological analyses failing to produce a consensus phylogeny. In this study, we use over 350 ultraconserved element (UCEs) loci comprising five million base pairs collected across thirty-five representative ostariophysan species to compile one of the most data-rich phylogenies of fishes to date. We use these data to infer higher-level (interordinal) relationships among ostariophysan fishes, focusing on the monophyly of the Characiformes— one the most contentiously debated groups in fish systematics. As with most previous molecular studies, we recover a non-monophyletic Characiformes with the two monophyletic suborders, Citharinoidei and Characoidei, more closely related to other ostariophysan clades than to each other. We also explore incongruence between results from different UCE datasets, issues of orthology, and the use of morphological characters in combination with our molecular data.
alignment-data.tar.gz
Alignment data for concatenated and coalescent-based analyses. Contents in README.
all-contigs-by-taxon.tar.gz
All contigs assembled by Trinity for each taxon in the data set. FASTA format by taxon. Contents in README.
astral.tar.gz
Gene trees and results from coalescent-based (ASTRAL) analyses. Contents in README.
bayesian-analyses.tar.gz
All files (samples, parameter estimates, convergence metrics, etc.) from Bayesian analyses. Contents in README.
fish-uce-1k-probes.tar.gz
Probe set used for UCE enrichment.
ml-analyses.tar.gz
All files (best trees, bootreps) from maximum likelihood analyses. Contents in README.
taxon-set.txt
The taxon set file for use with PHYLUCE.
tip-name-remappings.txt
A text file for PHYLUCE that we used to rename tips in each of the trees to fix taxonomy issues, incorrect spellings, etc.
trees.tar.gz
All trees generated for this manuscript. In newick format. Contents in README.
uce-contigs-by-taxon.tar.gz
All UCE contigs identified for each taxon. FASTA format by taxon. Contents in README.
processing-steps.txt
Text file with step-by-step description of data processing.
TableS1.xlsx
Supplemental Table 1: Number of reads, total number of base pairs sequenced, mean length of adapter- and quality-trimmed reads, 95 CI of trimmed read length, minimum read length, maximum read length, and median length for reads sequenced, by platform, from ostariophysan libraries enriched for UCEs (table also includes outgroup taxa, indicated with an asterisk).
TableS2.xlsx
Supplemental Table 2: Number of contigs assembled, total number of base pairs sequenced, mean contig length, 95 CI of contig length, minimum contig length, maximum contig length, median contig length, and number of contigs > 1 Kb in length from reads sequenced, by platform, from ostariophysan libraries enriched for UCEs (table also includes outgroup taxa, indicated with an asterisk).
TableS3.xlsx
Supplemental Table 3: Number of UCE loci enriched, total number of base pairs in all UCE loci, mean length of UCE loci, 95 CI of UCE locus length, minimum UCE locus length, maximum UCE locus length, median UCE locus length, number of UCE loci having a length > 1 Kb, UCE locus coverage, percentage of reads on-target, and percentage of on-target reads that were unique from reads assembled, by platform, from ostariophysan libraries enriched for UCEs (table also includes outgroup taxa, indicated with an asterisk). We collected Danio rerio UCE data from the danRer7 genome assembly.
Fig_S1.pdf
Supplemental Figure 1: Morphology based ML and Bayesian phylogeny. Support values on every node was (BPP >0.99 and bootstrap support of 100).
Fig_S2.pdf
Supplemental Figure 2: Results from ML analyses of ostariophysans using ultraconserved elements, with 50% complete data coverage, data not partitioned but concatenated.
Fig_S3.pdf
Supplemental Figure 3. Results from Bayesian analyses of ostariophysans using ultraconserved elements, with 50% complete data coverage, partitioned and concatenated.
Fig_S4.pdf
Supplemental Figure 4. Results from ML analyses of ostariophysans using ultraconserved elements, with 50% complete data coverage, partitioned and concatenated.
Fig_S5.pdf
Supplemental Figure 5: Results from ML analyses of ostariophysans using ultraconserved elements, with 75% complete data coverage, not partitioned but concatenated.
Fig_S6.pdf
Supplemental Figure 6: Results from ML and Bayesian analyses of ostariophysans using ultraconserved elements, with 75% complete data coverage, partitioned, and all data concatenated.
Fig_S7.pdf
Supplemental Figure 7: ASTRAL phylogeny of ostariophysans using ultraconserved elements with 50% data coverage, using a coalescent species tree approach. Internodes that were found to potentially be in the anomaly zone are shown with dashed lines (see text).
Fig_S8.pdf
Supplemental Figure 8: Result from a Bayesian analysis using ultraconserved elements (partitioned) and morphology, with 50% complete UCE data coverage.
Fig_S9.pdf
Supplemental Figure 9: Result from a ML analysis using ultraconserved elements (partitioned) and morphology, with 50% complete UCE data coverage.
Fig_S10.pdf
Supplemental Figure 10: Result from a Bayesian analysis using ultraconserved elements (partitioned) and morphology, with 75% complete UCE data coverage.
Fig_S11.pdf
Supplemental Figure 11: Result from a ML analysis using ultraconserved elements (partitioned) and morphology, with 75% complete UCE data coverage.