Hoplias maranhensis Guimarães, Rosso, González-Castro, do Nascimento Andrade, Brito, Guimarães, Díaz de Astarloa & Rodrigues, 2025

Guimarães, Karen L. A., Rosso, Juan J., González-Castro, Mariano, do Nascimento Andrade, Sarah J., Brito, Pâmella S., Guimarães, Erick C., Díaz de Astarloa, Juan M. & Rodrigues, Luís R. R., 2025, A new species of Hoplias (Characiformes, Erythrinidae) from the Maranhão State, western Northeast Atlantic basin, Brazil, Zoosystematics and Evolution 101 (4), pp. 1471-1493 : 1471-1493

Hoplias maranhensis Guimarães, Rosso, González-Castro, do Nascimento Andrade, Brito, Guimarães, Díaz de Astarloa & Rodrigues sp. nov.

Fig. 2 , Table 1

Hoplias malabaricus [non Bloch 1794]. Cardoso et al. (2018); Guimarães et al. (2022): ( Brazil, Maranhão, Itapecuru River; supported by distinct molecular methods as a putative new undescribed species in the H. malabaricus species-group).

Type materials.

Holotype. Brazil • UFOPA -I 001760 , 260 mm SL; Maranhão State, Itapecuru basin: Aldeias Altas Municipality, Alagadiço Stream ; coordinates: -4.55256, -43.32781; F. Lopes, 10 June 2024; BOLD: ACR 9466 .

Paratypes. All from Brazil, Maranhão State: Itapecuru Basin: Gonçalves Dias Municipality, Codozinho River : -5.11683, -44.09703, J. O. Sousa, 9 August 2023: • UNMDP 5375 , 1, 240 mm SL • UNMDP 5376 , 1, 258 mm SL • UNMDP 5377 (osteologic), 1, 280 mm SL • Caxias Municipality, Itapecuru River : -4,870 50, -43,361 16, A. Lima, 3 December 2021: • UFOPA -I 001757 , 4, 161– 325 mm SL • UNMDP 5393 , 1, 195 mm SL • UNMDP 5378 , 1, 180 mm SL • UNMDP 5394 , 1,198 SL • Cajazeira Stream : -4.99417, -43.47751, K. Guimarães, 5 February 2023: • UNMDP 5379 , 1, 184 mm SL • UNMDP 5380 (osteologic), 1, 116 mm SL • UFOPA -I 001761 , 1, 155 mm SL • Itapecuruzinho Stream : -4,926 69, -43,352 62, F. Lopes, 10 October 2023: • UNMDP 5381-5384 , 4, 138-180 mm SL • Ouro Stream : -4.81575, -43.25931, F. Lopes, 9 June 2024: • UFOPA -I 001758 , 3, 178– 212 mm SL ; Aldeias Altas Municipality, Alagadiço Stream : -4.55256, -43.32781, F. Lopes, 10 June 2024: • UFOPA -I 001759 , 2, 222– 240 mm SL ; Gonçalves Dias Municipality, Codozinho River : -5.11683, -44.09703, J. O. Sousa, 9 August 2023: • UFOPA / LGBIO–GCD 1 , 1, 225 mm SL • UFOPA -I 001762 , 1, 258 mm SL .

Non-type specimens.

All from Brazil, Maranhão State: Itapecuru Basin: Gonçalves Dias Municipality, Codozinho River : -5.11683, -44.09703, J. O. Sousa, 9 August 2023: • UFOPA / LGBIO–GCD 3 , 1, 204 mm SL • UFOPA / LGBIO–GCD 4 , 1, 195 mm SL • UFOPA / LGBIO–GCD 8 , 1, 283 mm SL ; Caxias Municipality, Itapecuru River: Cajazeira Stream : -4.99417, -43.47751, K. Guimarães, 5 February 2023: • UFOPA / LGBIO–CAX 8 , 1, 100 mm SL • UFOPA / LGBIO–CAX 9 , 1, 115 mm SL • UFOPA / LGBIO–CAX 10 , 1, 122 mm SL • UFOPA / LGBIO–CAX 11 , 1, 95 mm SL • UFOPA / LGBIO–CAX 12 , 1, 130 mm SL • UFOPA / LGBIO–CAX 13 , 1, 120 mm SL • UFOPA / LGBIO–CAX 14 , 1, 99 mm SL • UFOPA / LGBIO–CAX 15 , 1, 102 mm SL • UFOPA / LGBIO–CAX 16 , 1, 113 mm SL • UFOPA / LGBIO–CAX 19 , 1, 80 mm SL • UFOPA / LGBIO–CAX 23 , 1, 156 mm SL ; Peritoró Municipality, Preitoró River : -4,514 78, -44,319 54, E. C. Guimarães & P. S. Brito, 4 September 2021: • UFOPA / LGBIO–MPI 005 , 1, 61 mm SL ; MEARIM BASIN: Formosa da Serra Negra Municipality, Ribeirão Tamboril : -6.42949, -45.99744, E. C. Guimarães & P. S. Brito, 7 September 2021: • UFOPA / LGBIO–MPI 013 H 1, 1, 51.45 mm SL .

Diagnosis.

Hoplias maranhensis is distinguished from other species of the H. malabaricus species-group by the following combination of characters: 14–16 predorsal scales, 37–40 lateral-line scales, 39–40 vertebrae, 15 branched caudal-fin rays, ii-iii unbranched anal-fin rays, a straight line formed by the last vertical series of scales at the base of the caudal-fin rays, a marked ridge in the posterior angular end of the ascending process of premaxilla, postcleithrum 1 with an anterior expansion and in contact or nearly in contact with an elongated postcleithrum 2, the first proximal pterygiophore of the dorsal fin widely bifurcated, the number of principal caudal-fin rays articulating with lower hypural bones, foramen not in contact with premaxilla border and ascending process of the premaxilla expanded medially and with a dorsoposterior tip.

The number of predorsal scales (14–16), lateral-line scales (37–40), and vertebrae (39–40) distinguishes H. maranhensis from H. argentinensis (17–19, 41–44, and 42–43, respectively). H. maranhensis can be distinguished from H. microlepis by predorsal scales (14–16 vs. 17–19), lateral-line scales (43–46), and vertebrae (42–43). Pre-dorsal scales (14–16 vs. 18), lateral-line scales (37–40 vs. 40–41), and vertebrae (39–40 vs. 42) differentiate H. maranhensis from H. teres . The count of lateral-line scales (37–40) and vertebrae (39–40) distinguishes H. maranhensis from H. mbigua (41–44 and 42, respectively). H. maranhensis is distinguished from H. auri by the number of branched caudal-fin rays (15 vs. 12–15), unbranched anal-fin rays (iii – iv vs. ii), lateral-line scales (37–40 vs. 39–40), and vertebrae (39–40 vs. 38–39). Hoplias maranhensis differs from H. misionera by having a straight line formed by the last vertical series of scales at the base of the caudal-fin rays (vs. curved) and two vertical rows of scales on caudal fin (vs. four). The shape of the ascending process of the premaxilla distinguishes H. maranhensis from H. malabaricus (with a marked ridge vs. poorly developed) (Fig. 3 ). H. maranhensis also differs from H. auri and H. argentinensis in some osteological traits: from H. auri it differs in postcleithrum 1 (with an anterior expansion vs. no anterior expansion), postcleithrum 2 (more elongated and positioned closer to the base of postcleithrum 1 vs. shorter and more distinctly separated from postcleithrum 1), the first two proximal pterygiophores of the dorsal fin (a bifurcation with a widely divergent “V” shape vs. a parallel bifurcation, with a less separation between branches), articulation pattern of principal caudal-fin rays with lower hypural bones (one ray with parahypural, five rays with hypural 1, one ray with hypural 2 vs. two rays with parahypural, three rays with hypural 1 and two rays with hypural 2) (Fig. 4 ), foramen not in contact with premaxilla border vs. foramen in contact; dorsoposterior margin of premaxilla projecting medially with an ascending and pointed process vs. not projecting medially and lacking ascending and pointed process) (Fig. 5 ).

From H. argentinensis it differs in postcleithrum 2 (round-shaped and positioned closer to the base of postcleithrum 1 vs. a triangular-shaped and more distinctly separated from postcleithrum 1), the first two proximal pterygiophores of the dorsal fin (a bifurcation with a widely divergent “ V ” shape vs. a parallel bifurcation), and articulation pattern of principal caudal-fin rays with lower hypural bones (one ray articulates with parahypural, five rays with hypural 1, one ray with hypural 2 vs. two rays with parahypural, three or four rays with hypural 1, and one or two rays with hypural 2) (Fig. 3 ).

Compared to H. mbigua , H. maranhensis exhibits a straight or slightly convex dorsal head profile (vs. markedly concave) and a longer interorbital width (22.8–30.7 %, mean 27.2 % vs. 18.5–28.6 %, mean 24.3 %). Morphometric differences also separate H. maranhensis from other species in the group. A longer dorsal-fin base (17.3–21 %, mean 19 %, vs. 16.2–17.7 %, mean 16.9 %), a narrower snout (snout width 20.4–25.7 %, mean 23 %, vs. 29.4–29.5 %, mean 29.4 %), greater body depth (20.1–24.4 %, mean 22.1 %, vs. 17–20.6 %, mean 18.8 %), and greater upper jaw length 54.4 (50.1–58 %, mean 54.4 %, vs. 50.5–51 %, mean 50.7 %) distinguish H. maranhensis from H. teres . Hoplias maranhensis differs from H. microlepis in snout width (20.4–25.7 %, mean 23 % vs. 23.6–28.6 %, mean 26.8 %), pre-nasal length (11.8–18.5 %, mean 16.0 % vs. 15.4–23.8 %, mean 21.4 %), pre-dorsal length (45–50 %, mean 47.8 % vs. 47.7–52.2 %, mean 49.8 %), and upper jaw length (50.1–58 %, mean 54.4 % vs. 47.7–53.4 %, mean 50.1 %). Compared to H. malabaricus , H. maranhensis differs in having a shorter predorsal length (45–50 %, mean 47.8 %, vs. 48.4–54.4 %, mean 50.8 %), a longer dorsal-fin (29.5–34.1 %, mean 32.0 % vs. 28.9–31.1 %, mean 32 %), a longer dorsal-fin base (17.3–21 %, mean 19.0 % vs. 15.3–19.2 %, mean 17.6 %), a longer pre-nasal length (11.8–18.5 %, mean 16 % vs. 12.5–16.2 %, mean 13.8 %), and a shorter pre-pelvic length (48.5–55.7 %, mean 53 % vs. 52.1–58.1 %, mean 54.3 %). H. maranhensis also differs from H. auri by a greater body depth (20.1–24.4 %, mean 22.1 %, vs. 16.2–23 %, mean 19.9 %), a longer dorsal-fin base (17.3–21 %, mean 19.0 % vs. 15.1–19.5 %, mean 17.5 %), and a shorter pre-pelvic length (48.5–55.7 %, mean 53 % vs. 53.5–58.8 %, mean 55.7 %) (Fig. 6 ).

Description.

Morphometric data are summarized in Table 1 .

Body cylindrical, dorsal profile of head straight or slightly convex. Anterior profile of the head angular to slightly rounded in lateral view, with the greatest body depth at dorsal-fin origin. Dorsal surface of the head with a pronounced ridge at the posterior angular end of the ascending process of the premaxilla, in contact with the mesethmoid medially and the nasal bone laterally (see Fig. 2 ).

The medial margins of the contralateral dentaries converge at the midline in a V-shaped angle. Both upper and lower lips, fleshy. Anterior nostrils with an incomplete tubular skin flap partially covering the opening, while posterior nostril without fleshy flap and equidistant from the anterior nostril and anterior bony margin of the orbit. Infraorbitals 3 and 4 completely excluded from orbital ring.

Teeth caniniform in both jaws. Premaxilla with a single tooth row containing 9 (1), 10 (23 *), or 11 (1) teeth. First two medial teeth large, followed by three to five smaller teeth, ending with two larger canines. First and last canines in this series, the largest. One or two small teeth posterior to last large premaxillary canine, nearly in contact with the first small maxillary tooth. Maxillary row with four to six anterior teeth, progressively increasing in size, followed by 37–46 smaller teeth. External dentary series with one symphyseal tooth, followed by two smaller teeth, another large tooth equal in size to the symphyseal tooth, and largest dentary canine. This is succeeded by three to seven small teeth and a series of six to 11 teeth arranged in a repetitive pattern of one large tooth followed by one, two, or three smaller conical teeth. Internal dentary series with 18–19 very small conical teeth, positioned either immediately posterior or slightly anterior to the last tooth of the external series. Accessory ectopterygoid, a single bone, bearing 10 (2), 11 (2), 12 (7), 13 (6 *), 14 (3), or 15 (1) conical teeth along ventrolateral margin.

Dorsal-fin origin positioned at midbody. Dorsal-fin rays ii- 11 (4), ii- 12 (18 *), or ii- 13 (3). Anal-fin rays ii- 8 in all specimens. Pectoral-fin rays i- 12 in all specimens. Pelvic-fin rays i- 7 in all specimens. Total number of caudal-fin rays 17 (i- 15 - i, n = 23). Pectoral-fin tip separated from the pelvic-fin origin by 3–4 scales, while the pelvic-fin tip separated from vertical line through the anus by 2–5 scales.

Predorsal scales 14 (7 *), 15 (14), or 16 (1), arranged in irregular series. Lateral line with 37 (3), 38 (10), 39 (9 *), or 40 (1) perforated scales, with 1 (6) or 2 (15 *) unperforated scales beneath opercular membrane. Two vertical rows present on the caudal-fin rays, similar in size and shape to body scales. Longitudinal scale series between the dorsal-fin origin and the lateral line 5, while those between lateral line and pelvic-fin origin 4 to 5. Caudal peduncle encircled by 20 longitudinal scale series.

First epibranchial with 9–12 gill rakers, first raker somewhat elongated. One angular raker at cartilage. Ceratobranchial with four to five elongated and 10–12 plate-like denticulate rakers. Laterosensory canal along ventral surface of the dentary with four pores, some specimens displaying five pores on one side. Laterosensory canal on preopercle with six pores, while canal along the infraorbitals with 11 pores, distributed as follows: infraorbital 1: 3–4 pores; infraorbital 2: 1–3 pores; infraorbital 3: 1–2 pores; infraorbital 4: 1 pore; infraorbital 5: no pores; infraorbital 6: 3 pores.

Laterosensory system on dorsal surface of the head with 11 pores, distributed as follows: nasal: 2 pores; frontal: 4-5 pores; pterotic: 2 pores; and one pore between the parietals at the posterior end of their suture. Supraopercle and extrascapular with following combination of pores: 0–2 (10) or 1–1 (14 *). Total vertebrae 39 (n = 6) to 40 (n = 2).

Color in alcohol.

Background coloration predominantly black to dark brown, darker along dorsum and close to lateral line. Ventral surface white to pale yellowish, occasionally light brown in some specimens. Scales cycloid, with those on the dorsal half of body with dark-brown melanophores. Scales on ventral half with a dark vertical blotch on their anterior margins; in paler individuals blotch occupies half of scale surface.

Head with variable color patterns: in some specimens, entire head marbled, dark in others. Four distinct dark stripes radiating posteriorly from eye, aligned with infraorbitals 2, 3, and 6, as well as between infraorbitals 4 and 5, with latter the widest. In specimens ≤ 155 mm SL, these stripes are more evident, becoming progressively less conspicuous or almost imperceptible in larger individuals (Fig. 7 ).

A broad posterodorsal-dark blotch at the junction of opercle and subopercle. Chevron-shaped lateral blotches with vertex directed anteriorly and open arms posteriorly oriented, irregularly spaced, with distance between successive blotches decreasing toward caudal peduncle. These marks less conspicuous or even absent in larger specimens. A dark elliptical or slightly rounded spot on dorsal region of the caudal peduncle, often in contact with dorsal anteroventral branch of last chevron blotch. In specimens <150 mm SL, dark longitudinal stripe runs along lateral-line scales, covering approximately half of scale series immediately above and below the lateral line. In larger specimens, this stripe becomes indistinct or imperceptible.

Latero-ventral surface of dentaries with transverse brown bands or blotches, occasionally extending to maxilla and lips, or entirely unmarked. Fins light-cream to dark brown, with dark spots across rays and interaxial membranes. Anal fin characterized by larger spots forming irregular dark bands. Pelvic and pectoral fins with a yellowish hue, particularly noticeable in smaller individuals (<150 mm SL). In larger specimens, these fins are darker or with scattered spots across the rays and membranes.

Landmarks-based morphometric analysis.

The 23 normalized interlandmark distances, which were analyzed by PCA of the correlation matrix, produced six eigenvalues greater than 1 (5.805, 4.185, 2.831, 1.734, 1.494, and 1.098). The first three PCs explained more than 55 % of the variance in the data. Only correlations (between variables and components) higher than 0.5 were taken as significant (Table 2 ). The first three axes (PC 1, PC 2, and PC 3) of the PCA based on IlD allowed graphic segregation of the five species-groups analyzed. The multivariate ordination provided by the combination of the first two principal components (PC 1 – PC 2, Fig. 8 A, B ) allowed graphic segregation of H. misionera , H. auri , H. mbigua , and H. argentinensis (first, second, third, and fourth quadrants, respectively) almost without overlapping between them, being H. maranhensis located basically in quadrants I and II (PC 1 – PC 2, Fig. 8 A, B ). However, the PC 1 – PC 3 (Fig. 8 C, D ) and PC 2 – PC 3 (Fig. 8 E, F ) axes allowed the graphic segregation of H. maranhensis from the remaining four species. In this sense, H. maranhensis showed the highest loading for the ILD 3–5 (distance between the anteriormost extreme of interopercle and the origin of pectoral fin). It also displayed the lowest values for ILDs 1–3, 2–3 (variables that indicate the interopercle is located nearer to the tip of the snout), and 9–10, 7–10, 8–10 (which denote a smaller body at the level of the fourth box-truss) (PC 1 – PC 3, Fig. 8 C, D ; PC 2 – PC 3, Fig. 8 E, F ). In a minor way, H. maranhensis showed lower values for ILDs 3–4, 3–6, 4–5, 4–6, 5–6, 5–8, 6–7, and 7–8 (PC 1 – PC 2, Fig. 8 A, B ; PC 1 – PC 3, Fig. 8 C, D ; PC 1 – PC 4, Fig. 8 G, H ), denoting smaller box-trusses II and III, in particular when compared with H. misionera and H. argentinensis .

H. misionera was basically located on the first quadrant, denoting highest loadings for the 3–4, 4–5, 1–3, 1–4, and 2–4 IlDs (PC 1 - PC 2, Fig. 8 ) and higher loadings for the 2–3 and 1–2 IlDs (PC 2 – PC 3, Fig. 8 E, F ). All these ILDs denote a bigger head displayed by H. misionera specimens when compared with the other four species-groups analyzed. On the other hand, H. argentinensis was located in the fourth quadrant. It was characterized by highest loadings for the 3–6, 5–6, 5–8, 6–7, and 7–8 IlDs (PC 1 – PC 2, Fig. 8 A, B ), which evidence a robust body at the level of box-trusses II and III, and lower values for the 1–2, 1–4, and 2–4 IlDs (PC 2 – PC 3, Fig. 8 E, F ).

H. auri was basically located in the second quadrant (PC 1 – PC 2, Fig. 8 A, B ), being characterized by the highest loadings for 1–2 ILDs, higher loadings for the 1–3, 1–4, 2–3, and 2–4 ILDs, but also 3–4, and 4–5 (in a minor way). These variables represent the head shape, and in particular, 1–2 is in fact correlated to the snout, denoting that H. auri possesses a long snout when compared with H. mbigua , H. argentinensis , and H. maranhensis . Moreover, H. auri presented lower loadings for the 3–6, 5–6, and 4–6 (minor way) IlDs (variables that represent the relationship between the posterior part of the head and the origin of the dorsal fin) but also lower values for 5–8, 6–7, and 7–8 IlDs (PC 1 – PC 2 Fig. 8 A, B and PC 1 – PC 3 Fig. 8 C, D ) (belonging to the third box-truss and related to the dorsal-fin base and origins of pectoral and ventral fins) (Fig. 1 ).

Finally, H. mbigua showed the highest loadings for the 7–10 and 8–10 IlDs (which represent both the distances between the insertion of the dorsal and anal fins and the first dorsal caudal fin ray insertion) (PC 1 – PC 3, Fig. 8 C, D ; PC 2 – PC 3, Fig. 8 E, F ). This species also showed the lowest values for the 3–4 and 4–5 IlDs and lower values for 1–2, 1–3, 2–3, 1–4, and 2–4 but also 5–6, 5–8, 6–7, and 7–8 (PC 1 – PC 2, Fig. 8 A, B ).

The data corresponding to the 23 PCs of the PCA were employed to perform the DA. The DA for the 86 individuals belonging to the five species analyzed of Hoplias produced four significant canonical discrimination functions, where the first two explained 73.8 % of the total variance in the data (Wilks lambda = 0.001, P <0.000). Five groups were clearly defined, according to those defined a priori by the PCA, and their centroids and individuals were separated on both the first and second discriminant functions (Fig. 9 ). The DA correctly classified 100 % of the Hoplias individuals according to the species-groups defined a priori, whereas the cross-validated analysis correctly classified 93.0 % of the fish according to their body shape (Table 3 ). Accordingly, group misclassifications were scarce, being H. maranhensis misclassified as H. mbigua , H. argentinensis , and H. auri (6.3 % in each species-group; Table 3 ).

Osteological features.

Premaxilla: bone located at the anteriormost part of the skull, with a dorsal process folded posteroventrally, dorsally narrow and ventrally expanded, bearing teeth along the ventral margin. Bounded posteriorly by the mesethmoid and nasals, laterally by the maxilla, and contacting its counterpart at the snout tip. Approximately triangular in lateral view. A foramen medially, close to but not in contact with premaxilla border. Dorsoposterior margin of premaxilla projecting medially with an ascending and pointed process, accompanied by a marked ridge at its posterior angular (see Fig. 5 ).

Caudal skeleton.

The parahypural does not fully articulate with the hemal spine 38 or 39. The hypural 1 not fully aligned with the length of the parahypural, leaving a small distal gap. The hypural 2 articulates with the hypural 1 but does not extend completely along its length. It does not make contact with the parahypural but may connect anteriorly with the hypural 3. The hypural 3 is the only element articulating with the posterior face of the last centrum and makes full contact with the hypural 4. The hypural 4, in turn, connects with both the urostyle and entirely the hypural 5. The hypural 5 exhibits full contact along its lower margin with the hypural 4, while its upper margin articulates completely with the urostyle and the hypural 6. The hypural 6 fully connects to the hypural 5 along its lower margin and dorsally to both the urostyle and the uroneural bones. The uroneural exhibits full contact with the Urostyle along its upper margin and may also contact the hypural 6 along its lower margin. The epural is entirely in contact with the urostyle and the hemal spine 38 or 39 in its middle region. Hypural bones receive nine principal caudal-fin rays dorsally and eight ventrally. The rays articulate with the hypural bones as follows: the lower unbranched ray articulates with hemal spine 38 or 39; first branched ray is associated with the parahypural; next five branched rays articulate with hypural 1; one with hypural 2; two to three with hypural 3; two to three with hypural 4; two with hypural 5; and one with hypural 6. The latter may be either the last branched ray or the upper unbranched ray, which can articulate with either hypural 6 or the uroneural (Fig. 10 ).

Pectoral girdle.

Pectoral girdle is composed of the posttemporal, supracleithrum, postcleithrum 1–3, mesocoracoid, scapula, cleithrum, and coracoid bones. Posttemporal bifurcated, with a dorsoventrally flattened ascending branch, its anterior portion resembling an arched blade connected to the dorsal surface of the skull. The descending branch, thin and elongated, rod-like in shape. Supracleithrum form the upper portion of the pectoral girdle, an elongated structure, slightly curved at the apex, dorsally articulated with the posttemporal. Ventrally, slightly wider, connected to the cleithrum. Postcleithrum 1, a rounded bone positioned medially to the supracleithrum and laterally to the cleithrum, with a slight anterior expansion. Dorsally located to the ventral end of the supracleithrum. Postcleithrum 2, smaller and laterally narrower than postcleithrum 1, elongated in shape and positioned closer to the base of postcleithrum 1. Connected to the posterior process of the cleithrum and to postcleithrum 3. Postcleithrum 3, slender and narrow, rod-like, positioned below postcleithrum 2, extending to the horizontal line passing through the ventral margin of the coracoid. Scapula a dorsally bifurcated structure at the base of the cleithrum, articulated with the first ray of the pectoral fin. Mesocoracoid a thin, curved bone between the cleithrum and the scapula. Cleithrum dorsally articulated with the supracleithrum. An arched bone, broader in the lower portion, gradually narrowing into a curved tip at the apex, directed toward postcleithrum 1. Coracoid anteriorly articulated with the cleithrum. A bone with a triangular outline, a thick base, and a tapered end. Fully connected to the cleithrum, with one foramen in the medial portion (Fig. 11 ).

Dorsal-fin.

A series of 13 distal pterygiophores and 14 proximal pterygiophores. First two proximal pterygiophores fused, and only the second one in contact with a distal pterygiophore. The first proximal pterygiophore (lacking a distal pterygiophore) directly receives the first unbranched dorsal-fin ray. The second proximal pterygiophore in contact with the first distal pterygiophore, and the latter receiving the second unbranched dorsal-fin ray. Distal pterygiophores small, positioned at the junction between the lower end of the fin rays and the proximal pterygiophores. The first two fused proximal pterygiophores, with an anterior bifurcation in a widely divergent “ V ” shape. Remaining proximal pterygiophores simple, slenderer, and progressively smaller posteriorly (Fig. 12 ).

Molecular species delimitation.

Molecular analyses using the COI gene confirmed the distinctiveness of the new species within the H. malabaricus species-group. Species delimitation methods consistently recognized H. maranhensis as a genetically distinct lineage. The Barcode Index Number (BIN) algorithm assigned a unique BIN tag (BOLD: ACR 9466), indicating clear separation from other Hoplias species clusters. Similarly, Automatic Barcode Gap Discovery (ABGD) and Assemble Species by Automatic Partitioning (ASAP) analyses identified a distinct operational taxonomic unit (OTU) for the species, supporting its independent lineage status within the genus (Fig. 13 ).

Pairwise genetic divergences between H. maranhensis and related taxa revealed its closest genetic affinity with H. malabaricus (BOLD: AFU 2064) and H. auri (BOLD: ADL 3159), both exhibiting 3.0 % divergence. This was followed by H. mbigua (BOLD: AAI 8239) at 6.0 % divergence and H. microlepis (BOLD: AAD 3629) at 7.0 %. Within the H. malabaricus species-group, the most divergent taxa from H. maranhensis sp. n. were H. misionera (BOLD: AAB 1732) and H. argentinensis (BOLD: AAZ 3734), both with 9.0 % divergence. The greatest overall genetic distance was observed between H. maranhensis sp. n. and the outgroup H. lacerdae (BOLD: ABW 2258), with a divergence of 17.0 % (Table 4 ). Phylogenetic reconstruction using a Neighbor-Joining (NJ) tree based on Kimura two-parameter (K 2 P) distances supported these findings, delineating the new species as a well-supported monophyletic clade (Fig. 13 ).

Etymology.

The specific epithet maranhensis is related to the place of collection: Maranhão State, in the northeastern, western Atlantic drainages of Brazil.

Distribution.

The new species is known from the Western Northeast Atlantic hydrographic region of Maranhão State, predominantly distributed throughout the Itapecuru River basin, including the main river and streams. It also occurs in right-bank tributaries of the Mearim River drainage (Fig. 14 ).

Conservation issues.

The newly described species inhabits river systems in Maranhão State, a region increasingly affected by anthropogenic disturbances, including deforestation, agricultural expansion, and urbanization. These activities contribute to habitat degradation through increased sedimentation and water pollution from agrochemicals and untreated domestic sewage ( Bragança 2018, Mataveli et al. 2021, Marengo et al. 2022, Santos et al. 2024). Furthermore, environmental changes can lead to reduced habitat availability, changes in trophic dynamics, and overall declines in freshwater biodiversity ( Eastwood et al. 2023). Although trahiras are commonly targeted in subsistence, commercial, and aquaculture in Maranhão State ( Guimarães et al. 2021 c), the fishing pressure for the fish species is unquantified ( Santos et al. 2023). Overfishing, particularly in systems already experiencing environmental degradation, could further threaten population stability. Given these factors, continuous monitoring of habitat integrity, assessment of potential fishery impacts, and implementation of conservation strategies are essential to mitigate eventual threats and ensure the long-term persistence of H. maranhensis . Moreover, our sampling program directed to collect specimens of the new species allowed us to record the occurrence of H. malabaricus living in sympatry with H. maranhensis . This fact poses an additional challenge for any management or conservation strategy intended to be specifically applied to any one of these very similar species.

Figures

Figure 1. Box-truss showing interlandmark distances based on 11 homologous anatomical landmarks collected in the specimens of the five different species of the Hoplias malabaricus species-group analyzed. Box-truss numbers in Roman numerals.

Figure 2. Hoplias maranhensis sp. nov., holotype, UFOPA - I 001760, 260 mm SL, Alagadiço Stream, Itapecuru River drainage, Brazil.

Figure 3. Comparative dorsal view of the head showing the pronounced ridge at the posterior angular end of the ascending process of the premaxilla in Hoplias maranhensis vs. slightly developed or no ridge in Hoplias malabaricus. H. maranhensis: UFOPA / LGBIO – CAX 12, 130 mm SL (A); UFOPA - I 001757, 161 mm SL (B); UNMDP 5393, 195 mm SL (C); and H. malabaricus: MNHN-IC- 1901-0381, 132 mm SL (collected in Ouanary River, French Guiana, in 1976) (D); MHNG 2752.085, 159 mm SL (collected in Saramacca River, Suriname, in 2014) (E); MNHN 2003-2527, 204 mm SL (collected in Orapu River, French Guiana, in 2003) (F).

Figure 4. Comparative osteological differences between Hoplias maranhensis UNMDP 5377, 280 mm SL (A, D, G); Hoplias auri UFOPA / LGBIO-CRP 71, 235 mm SL (B); UNMDP 5391, 211 mm SL (E); UFOPA / LGBIO-CRP 74, 246 mm SL (H); and Hoplias argentinensis UNMDP 5390, 287 mm SL (C, F, I). A – C. Pectoral girdle, showing the postcleithrum 1 (Pc 1) and postcleithrum 2 (Pc 2); D – F. First proximal pterygiophore (Pt 1) of the dorsal fin; G – I. Articulation pattern of principal caudal-fin rays with lower hypural bones, showing differences in ray connections with parahypural (Ph), hypural 1 (Hy 1), and hypural 2 (Hy 2).

Figure 5. Comparative morphology of the premaxilla in Hoplias maranhensis UNMDP 5377, 280 mm SL (A, C) and Hoplias auri UNMDP 5392, 215 mm SL (B, D). A, B. Frontal view of the skulls, showing the shape of the premaxillary bone (Pm); C, D. Illustration of the premaxilla, highlighting differences in the shape of the ascending process, the medial expansion, and the foramen position.

Figure 6. Box plots comparing measurements that differentiate species of the H. malabaricus species-group in freshwater of Brazil: Hoplias auri (blue), Hoplias malabaricus (pink), and Hoplias maranhensis (orange). Each box represents the interquartile range, with the horizontal line indicating the median. Whiskers extend to the minimum and maximum values, and outliers are shown as individual points. The red cross marks the mean value for each species.

Figure 7. Progressive reduction of dark stripes radiating from the infraorbitals in Hoplias maranhensis. A. UNMDP 5380, 116 mm SL; B. UFOPA / LGBIO – CAX 17, 155 mm SL; C. UFOPA - I 001757, 170 mm SL. Scale bar: 10 mm.

Figure 8. Principal Component Analysis (PCA) of the five species of the Hoplias malabaricus species-group. Left: Projections of individuals’ scores onto the fourth factorial plane of PCA; PC 1 - PC 2 (A); PC 1 - PC 3 (C); PC 2 - PC 3 (E); PC 1 - PC 4 (G). Right: Correlation between interlandmark distances and principal components: PC 1 - PC 2 (B); PC 1 - PC 3 (D); PC 2 - PC 3 (F); PC 1 - PC 4 (H). H. mbigua (black squares), H. misionera (green diamonds), H. argentinensis (blue diamonds), H. auri (yellow triangles), H. maranhensis (red triangles).

Figure 9. Discriminant analysis of the five species of the Hoplias malabaricus species-group: H. maranhensis (red triangles), H. auri (yellow triangles), H. argentinensis (blue diamonds), H. mbigua (black squares), and H. misionera (green triangles). Black empty squares indicate group centroid location.

Figure 10. Caudal skeleton illustration of Hoplias maranhensis, showing the articulation pattern of principal caudal-fin rays with hypural bones.

Figure 11. Pectoral girdle illustration of Hoplias maranhensis, showing the shape and spatial arrangement of bones, including the anterior expansion of postcleithrum 1, an evaluated informative trait.

Figure 12. Illustrative dorsal-fin structure of Hoplias maranhensis.

Figure 13. Neighbor-joining tree of Hoplias maranhensis and related species of the H. malabaricus species-group. Blue indicates the new species. The black lateral bars show the partitions of species delimitation through BIN, ABGD, and ASAP analyses. Values in branches indicate bootstrap values.

Figure 14. Known distribution of Hoplias maranhensis in the Western Northeast Atlantic based on material examined herein (circles indicate paratype localities; star indicates type locality).