S1 Text. Summary of cytogenetic observations and mechanisms for unidirectional introgression.
(DOCX)
S1 Table. Sample information, BioSample IDs, metadata, ploidy inference and filtering.
(XLSX)
S2 Table. Summary statistics for each of the species calculated with R package “PopGenome”.
(XLSX)
S3 Table. The average test AUC (area under the Receiving Operator Curve) for the replicate
runs for all the species in MaxEnt modeling for predicting species distribution from
climate data at the species occurrences.
(XLSX)
S4 Table. Instances of polyploid Neobatrachus. (XLSX)
S1 Fig. Species tree and admixture results for optimal clustering at K equals 3, 7 and 9 (see
S4 Fig. for optimal number of clusters). Vertical colored bars to the left of the tips of the tree correspond to our final species assignments (S1 Table); colors of the bars are species-specific
and correspond to the branch colors from Fig 1A; filtered out samples are marked with black
bars.
S2 Fig. Nuclear species tree as inferred using ASTRAL, all nuclear loci, and complete taxon
sampling. Figure extends across four parts (A, B, C, D) and is color coded by species identity.
S3 Fig. Two dimensional representations of MDS gene tree space, colored by optimal clustering
scheme for two dimensions (k = 2) and three dimensions (k = 4), and their associated
topologies inferred using ASTRAL. Each point represents a single gene tree, colored
clusters match colored trees displayed to the right. Nodes at values indicate bootstrap support.
S4 Fig. Cross-validation plot showing three local optimal solutions for ADMIXTURE clustering
at K equals 3, 7 and 9.
S5 Fig. (A) Gene trees, colored by clade, for 361 nuclear loci based on 2 individuals per species
show considerable incongruence and differ from the species trees (bold black topology). (B)
Gene trees for diploid individuals only also show considerable incongruence and differ from
the species trees (bold black topology). (C,D) Species tree colored by topological consistency as
measured by gene concordance factors—gCF%, the percentage of loci which decisively favor a
given bipartition. Warmer colors indicate high discordance, cooler colors indicate strong concordance.
S6 Fig. Genealogies for six randomly sampled nuclear loci (y-axis) with different diploid
individuals chosen as representatives for each species (different sample sets, x-axis) are
consistent with each other. Genealogical conflict remains only among loci. This supports a
scenario of rapid speciation of the diploid species without secondary contact or persistent
incomplete lineage sorting.
S7 Fig. Sequenced loci statistics on alignment length and number of variable sites inferred
by AMAS (11).
S8 Fig. Distribution of allele frequencies of biallelic sites in Neobatrachus tetraploids supports
tetrasomic inheritance mode in N. sudellae and N. aquilonius and mixed inheritance
mode in N. kunapalari. (A) Pairwise combination of individuals within the diploid species
model the expected allele frequencies in autotetraploids with tetrasomic inheritance (blue
line), when pairwise combination of individuals between the diploid Neobatrachus species
model the expected distribution for allotetraploids with disomic inheritance mode (purple
line). Modeled allotetraploids show excess of intermediate allele frequencies compared to autotetraploids.
Gray area shows 95% confidence interval. (B) Comparing the ratio between intermediate
(40–60%) and rare (<30%) allele frequencies we reject allotetraploid origin for N.
sudellae and N. aquilonius, when N. kunapalari shows intermediate distribution, suggesting
mixed inheritance. Comparisons performed with Wilcoxon tests adjusted for multiple testing.
S9 Fig. SnaQ analysis. A. The optimum phylogenetic network includes two hybridization
events. B. Network score has the best support at minumum 2 hybridization events, additional
allowed hybridizations do not increase the network score.
S10 Fig. Heatmap and hierarchical clustering of the Neobatrachus lineages based on the
distance matrix from pairwise median Fst values. Tetraploid species (N. sudellae, N. aquilonius
and N. kunapalari; highlighted with black left bar) cluster together and are characterised
by the lowest Fst values between each other. This, together with low Fst values between tetraploid
and diploid lineages, can probably be explained by the gene flow within the tetraploids
and between the diploids and the tetraploids. Diploid lineages (highlighted with grey left bar)
appear to be more isolated from each other compared to tetraploids, which is in agreement
with ADMIXTURE assignment results and TreeMix estimations of possible migration events.
S11 Fig. Occurrence data locations registered at the AmphibiaWeb database for Neobatrachus
species: A—tetraploids, B—diploids.
S12 Fig. PCA analysis of bioclimatic variables for Neobatrachus entries in the occurrence
AmphibiaWeb database. A) Barplot showing the percentage of variances explained by each
principal component. The first three principal components are labeled with the top three contributions
of variables. BIO10 = Mean Temperature of Warmest Quarter, BIO12 = Annual
Precipitation, BIO17 = Precipitation of Driest Quarter, BIO18 = Precipitation of Warmest
Quarter, BIO19 = Precipitation of Coldest Quarter. B-D) Pairwise combinations of the first
three principal components, where individuals with a similar profile of bioclimatic data are
grouped together. Points represent each individual and colored according to the species
assignment, ellipses represent 95% confidence area.
S13 Fig. The results of the jackknife test of variable importance for models on each species.
BIO19 (Precipitation of Coldest Quarter) was the most informative variable for the models of
N. pelobatoides and N. albipes distributions; BIO18 (Precipitation of Warmest Quarter) was
the most informative variable for the models of N. wilsmorei, N. sutor and N. kunapalari;
BIO17 (Precipitation of Driest Quarter) was the most informative variable for the model of N.
fulvus; BIO10 (Mean Temperature of Warmest Quarter) for N. pictus; and BIO9 (Mean Temperature
of Driest Quarter) for N. sudellae and N. aquilonius.
S14 Fig. The point-wise mean of the 10 models for each of the diploid species build on
environmental layers from the current climate data and applied to the environmental layers
from the Last Glacial Maximum climate data.
S15 Fig. The point-wise mean of the 10 models for each of the tetraploid species build on
environmental layers from the current climate data and applied to the environmental layers
from the Last Glacial Maximum climate data.
S16 Fig. Karyotypes of Neobatrachus. A) N. sutor [2n], B) N. pictus x N. sudellae triploid [3n]
hybrid from Moyston, east of the Grampians, Victoria, C) N. fulvus x N. sutor triploid [3n]
hybrid from Learmonth, Western Australia, D) N. sudellae [4n], E) tetraploid x tetraploid hybrid from north of Menzies, Western Australia, F) N. pictus x N. sudellae pentaploid [5n]
hybrid from Moyston, east of the Grampians, Victoria. Arrowheads indicate nucleolar organiser
regions (NORs).
