Fig 1.
Drainage basins in: (a) Levant (b) Iran (the Maharlu basin lies between the Rud-e Kor and the Gulf basins).
Drawing of Fig 1(b) reprinted from [9] under a CC BY license, with permission from author with modifications.
Table 1.
Accession numbers of COI sequences are written in bold and those of LSU sequences are written in italics.
Fig 2.
Condensed cladogram obtained from COI sequences using Maximum Parsimony/Bayesian analysis.
Numbers above branches refer to bootstrap/posterior probability percentages; only values ≥ 50% are shown. “-” indicates that no bootstrap value was obtained from MP analysis as only a single sequence was included in the analysis.
Fig 3.
Haplotype network for the Capoeta caelestis, Capoeta damascina, and C. umbla COI sequences showing the number of nucleotide differences between haplotypes.
Clades labeled A and B correspond to clades A and B in the phylogenetic tree.
Fig 4.
Haplotype network for the Capoeta buhsei, Capoeta coadi, and Capoeta saadii COI sequences.
Clades labeled C, D, and E correspond to clades C, D, and E in the phylogenetic tree. These clades are not linked to clades A and B as the number of nucleotide differences exceeds the chosen connection limit (10 mutation steps).
Fig 5.
Condensed cladogram obtained from LSU sequences using Maximum Parsimony/Bayesian analysis.
Numbers above branches refer to bootstrap/posterior probability percentages; only values ≥ 50% are shown. “-” indicates that no bootstrap value was obtained from MP analysis as only a single sequence was included in the analysis.
Fig 6.
Condensed cladogram obtained from COI+LSU sequences using Maximum Parsimony/Bayesian analysis.
Numbers above branches refer to bootstrap/posterior probability percentages; only values ≥ 50% are shown. “-” indicates that no bootstrap value was obtained from MP analysis as only a single sequence was included in the analysis.
Fig 7.
A plausible biogeographical scenario for the separation between the (a) eastern and (b) western lineages.