Testing species delimitations in four Italian sympatric leuciscine fishes in the Tiber River: a combined morphological and molecular approach

Abstract

Leuciscine fishes represent an important component of freshwater ichthyofauna endemic to northern Mediterranean areas. This lineage shows high intra-specific morphological variability and exhibits high levels of hybridization, two characteristics that contribute to systematic uncertainties, misclassification of taxa and, potentially, the mismanagement of biodiversity. This study focused on brook chub, Squalius lucumonis, an endemic taxon of Central Italy. The taxonomic status of this species has long been questioned, and a hybrid origin from sympatric leusciscines (S. squalus x Rutilus rubilio, or S. squalus x Telestes muticellus) has been hypothesised. A phenotypic (evaluating shape and meristic counts) and genetic (using mitochondrial and nuclear markers) investigation of these four taxa was conducted to test species delimitation in sympatric areas and to evaluate the taxonomic status of S. lucumonis. One hundred and forty-five individuals of all four taxa were collected within streams of the lowest portion of the Tiber River basin and analysed; this region encompasses a large portion of the S. lucumonis distribution. The different morphological and genetic approaches were individually examined, compared, and then combined in a quantitative model to both investigate the limits of each approach and to identify cases of misclassification. The results obtained confirm the cladogenetic non-hybrid origin of S. lucumonis, highlight the need for immediate conservation actions and emphasise the value of an integrated approach in the study of leuciscines evolution.

Figure 1. Landmarks used for Geometric morphometrics analysis of external shape.

(1) Snout tip; (2) nostril; (3) beginning of scales coverage on the dorsal outline; (4) centre of the eye; (5) posterior extremity of the premaxillar; (6) insertion of the operculum on the ventral lateral profile; (7) beginning of the lateral line; (8) posterior extremity of the operculum; (9) superior and (10) inferior insertions of the pectoral fin; (11) anterior and (12) posterior insertions of the dorsal fin; (13) superior and (17) inferior insertions of the caudal fin; (14) superior and (16) inferior insertion of the caudal peduncle: (15) posterior body extremity; (18) posterior and (19) anterior insertion of the anal fin; (20) anus; (21) anterior insertion of the pelvic fin.

Figure 2. Classification tree obtained from the CRT analysis.

NSLL: number of scales of the lateral line; NRAF : number of rays of the anal fin; NRPF : number of rays of the left pectoral fin. Rr: R. rubilio; Ss: S. squalus; Sl: S. lucumonis; Tm: T. muticellus.

Figure 3. CVA output of the overall comparisons of the four taxa; the pattern described by the first two discriminant axes is shown.

Convex hulls are used to delimitate groupings; splines illustrate shape changes along each axis. Rr: R. rubilio; Ss: S. squalus; Sl: S. lucumonis; Tm: T. muticellus. See text for arrowed individuals.

Figure 4. Species trees estimated with concatenation and coalescent approaches.

Panel A: ML and BI trees of concatenated dataset (RAG1 and cyt b sequences); bootstrap values (>70%) and Bayesian posterior probabilities (>0.7) are reported. Sl-Arno, Ss-Arno, Ss-Vipava, Tm-Tiber idicate specimens whose cyt b and RAG1 sequences were available in GenBank (see Table S4). Specimen Ss22 showing a S. lucumonis-like nuclear sequence, and Sl14 showing both mitochondrial and nuclear R. rubilio-like sequences, are boxed. Panel B: Phylogeny estimated with *BEAST; posterior probabilities shown for all nodes.

Figure 5. Scatterplot representing the pattern obtained by partial least square (PLS) analysis of shape (GM) and genetic (PCOA axes) data.

Splines show shapes relative to the extremes of the y-axes. Specimen Ss22 and Sl14 are indicated by arrows.