Eigenmann 1916 :84 Rendahl 1937 :2 Hemiancistrus furtivus H. furtivus Hemiancistrus H. landoni H. landoni H. hammarlundi Rendahl 1937 H. fugleri Ovchynnyk 1971 H. landoni H. annectens H. hammarlundi H. furtivus H. fugleri H. annectens H. landoni H. furtivus H. annectens H. furtivus H. fugleri H. furtivus Hemiancistrus Hemiancistrus H. landoni H. furtivus H. landoni Hemiancistrus annectens H. aspidolepis H. maracaiboensis H. holostictus H. wilsoni H. annectens Hypostomus Hypostomus Hemiancistrus Ancistrini Hypostomus annectens Hemiancistrus landoni H. annectens H. landoni H. furtivus Ancistrini H. annectens Hypostomus Hemiancistrus The species of Hemiancistrus (Siluriformes: Loricariidae) from Ecuador Provenzano, Francisco Barriga, Ramiro Zootaxa 2017 4272 2 221 235 6LNNX 1633055 Eigenmann 1916 Eigenmann 1916 [151,645,600,627] Actinopterygii Loricariidae Hemiancistrus Animalia Siluriformes 8 229 Chordata species landoni      Hemiancistrus landoni  Eigenmann 1916:84. Typelocality: Naranjito, Ecuador.  Hemiancistrus hammarlundi  Rendahl 1937:2, Fig. 1.  Typelocality: Río Clementinasystem, northwest of Babahoyo, Los Ríos, Ecuador.     Materialexamined. Allfrom Ecuador, El Oro Province, MEPN-5948, 1 ex., 171.5 mmSL, La Cuca, canal de riego, río Arenillas, GranjaPREDESUR, approx. 03°30'00"S 80°04'20"W, R. Barriga et al.,  13 April 1979. MEPN- 5954, 1 ex., 88.6 mmSL, Río Piedras, curso superior del río Arenillas, approx. 03°38'10"S 79°55'40"W, R. Barriga et al.,  12 April 1979. MEPN-17505, 1ex. 68.7 mmSL, Río Zaracay, approx. 03°38'55"S79°52'180" W, P. Tufiño& R. Barriga,  20 November 2010.  Guayas Province, MEPN-9820, 1 ex., 59.9 mmSL, Río Minas, donde cruza la cooperativa 23 de Noviembre,  9 kmS de Naranjal, approx., 02°41'26"S 79°38'16"W, R. Barriga,  22 September 1992. MEPN-15118, 1 ex., 134.9 mmSL, Río Minas, donde cruza la cooperativa 23 de Noviembre,  9 kmS de Naranjal, approx., 02°41'26"S 79°38'16"W, R. Barriga,  22 September 1992. Los Ríos Province, MEPN-9926, 8 ex. (7 Alc. 1 Dry skeleton), 128.4–205.8 mmSL, Río Catarama, cerca de la poblacion Corona, approx. 01°37'23"S 79°28'20"W, R. Barriga, 15 February 1997. MEPN-10389, 1 ex., 214.6 mmSL, Río Quevedo, approx. 01°02'04"S 79°27'43"W, C. Estrella,  May 1953. MEPN-17049, 1 ex., 60.6 mmSL, Río Jujan, cerca de la población de Jujan, approx. 01°53'04S 79°33'06"W, S. Abril,  23 June 1986. MEPN-17050, 1 ex., 84.8 mmSL, Quevedo, G.Onore,  November 1983. MEPN-18151, 3 ex., 192.3–213.4 mmSL, Río Las Juntas, afluente del río Babahoyo, approx. 01°52'06"S 79°22'43"W, M. Olalla, 12 March 1964.  Manabí Province, MEPN-9099, 1 ex., 135.6 mmSL, Río Portoviejo, Laguna de Poza Honda, parte inferior, approx., 01°04'55"S 80°09'16"W, R. Barriga, 0  5 February 1993. MEPN-17952, 6 ex., 56.3–183.9 mmSL, Río Portoviejo, 1 kmaguas abajo de la primera compuerta, approx., 01°06'21"S 80°09'41"W, R. Barriga et al.,0  3 October 1992.   Hemiancistrus landoniis recognized among all species of  Hemiancistrus sensu lato, except for  H. mediansby its abdomen, which is completely covered by small bony plates, and has a peculiar color pattern, rounded dark spots ( vs.abdomen naked with light spots or plain). According to the figures and redescription indicated by Fisch- Muller et al.(2012),  H. landonican be recognized from  H. medians, by its dorsal fin when folded does not reach the origin of the adipose fin versusthe dorsal fin folded reaches the origin of the adipose fin. Further recognized by its head shape, which is less deep, and by its supraoccipital without a keel. In  H. medians, the head is massive, and the supraoccipital has a keel. The eye is smaller in  H. landonithan in  H. medians, 15%–16% HL versus18%–27% HL, respectively. Finally, the number of cheek odontodes is smaller in  H. landonithan in  H. medians, 10–20 vs.20–60. Some large specimens show highly developed, long and pointed odontodes, on distal and dorsal portion of pectoral spines. This condition, as in other cases, may occur in sexually active adult males.  Hemiancistrus landoniis an endemic species of the Pacific slope of Ecuador, and is found exclusively in aquatic systems of the GuayasRiver Basin and other water courses related to this basin.   Comparative osteology.On the three species identified, we examined the bones that are related to the mobility of the opercular bone. The bones are the sphenotic, the compound pterotic (from now the pterotic), the hyomandibular and the opercular. The observed changes in the state or condition of each bone are related to the development of the musculature involved in the movement of the opercular and concomitant of the developed odontodes associated. The observed changes on the sphenotic occur on the posterior surface. Two conditions are perceived; in the first condition, the posterior region has an excavation or canal, extending from the middle of the bone to the ventral region of the bone; the area from the middle of the bone towards the dorsal region is massive ( Fig. 6A). The second condition is an excavation or canal that runs along the entire posterior surface, from the dorsal to the ventral. This channel is open in the ventral region, but towards the dorsal border it may be totally or partially open ( Fig. 6B). In  H. annectensthe first condition is observed, while in  H. landoniand in  H. furtivus, the second condition is observed, the canal is almost occluded at the dorsal border.   FIGURE 4.Lateral and ventral view of  Hemiancistrus landoni, MEPN 17952, 135.4 mm SL. The pterotic bone has on its anterior surface something similar to that indicated for the sphenotic. In the first condition we can observe the presence of a canal that is occluded or closed towards the dorsal region of the bone ( Fig. 7A). In the second condition the canal runs all or nearly all of the entire anterior surface of the pterotic bone ( Fig. 7B). By articulating the sphenotic and pterotic bones, the canal partially becomes a tunnel. The tunnel may be closed or open dorsally. In  H. annectens,the first condition is observed, the pterotic canal is occluded or the tunnel is closed. In  H. landoniand in  H. furtivus, the second condition is observed, here the channel and tunnel runs almost all the anterior surface of the pterotic and the posterior surface of the sphenotic. In this case the tunnel closes towards the dorsal region. On the ventral and anterior border of the pterotic there is a fossa and a process (anterior process of the pteroticsupracleitrum according to Armbruster 2004). Again, two conditions related to this region are observed. In the first condition, the fossa and the process are slightly developed ( Fig. 7A). In the second, there is an appreciable development of the fossa and the process ( Fig. 7B). In  H. annectens,the first condition is observed, while in  H. landoniand in  H. furtivus, the second condition is observed. On the analyzed species, the opercular bone presents two different forms ( Fig. 8). In the first condition, the general shape or contour of the bone is almost triangular, similar to that described by Schaefer (1987)for  Hypostomus plecostomusand  Cochlidon cochliodonand Armbruster (2004)for  Hemiancistrus maracaiboensisand  Hypostomus taphorni(condition 0). In the second condition, the opercular has the contour of the opercular in the form of a paddle with a posterior dorsal projection, similar to that described by Schaefer (1987)for  Peckoltia niveataand by Armbruster (2004)for  Peckoltiasp. (Condition 1, Type-  Peckoltia). The first condition of the opercular is present in  H. annectens( Fig. 8A), while the second condition is observed in  H. landoniand  H. furtivus( Fig. 8B). The changes in opercular shape can be related to the increase in its rotational capacity at the point of support (condyle of articulation with the hyomandibular) and its properties as a lever. The conditions or states observed could be used to include the species under study into the Hypostomini or Ancistrini sensu Armbruster (2004). The changes noted in the three bones are linked to increased development (increased diameter or muscle thickness) and increased surfaces of origin and insertion of the levator operculiand dilator operculimuscles, which was evidenced during the preparation of skeletons.   Discussion.The status of the genus  Hemiancistrusis still controversial. The loss of the holotypeof  H. medians, and the way Kner (1854)presented the original description, generated doubts and discrepancies about the correct identity of this species. Fisch-Muller et al.(2012) established the taxonomic status and the identity of the typespecies, in spite of the descriptions provided by Günther (1864)and Regan (1904). We do not have specimens of  H. medians, however, as indicated by Fisch-Muller et al.(2012) and Armbruster et al.(2015),  H. mediansand  H. landonishare some external morphological characteristics not observed in the other species included in  Hemiancistrus sensu lato, such as the presence of keels in the lateral plates of the body and a similar coloration pattern. The way the abdomen is covered appears to have some difference between  H. mediansand  H. landoni, but none of the other species included in the group has such an extended abdominal covering. According with Armbruster et al.(2015) only  H. mediansis included in the genus  Hemiancistrus. Therefore, the generic status of  H. landoniand the other species remains uncertain.    Hemiancistrus furtivusis represented at the MEPN by seven specimens in five cataloged lots. Only four are in good conditions, and/or with an appreciable standard length. Comparisons with original descriptions of the geographically nearest species were made carefully, and the external characters observed on  H. furtivusshows that it is a new species. The belly is totally naked, the dorsal and caudal fins have a distinctive color pattern, and the small movable cheek odontodes differ from that of other Pacific Coast species described or placed in  Hemiancistrus. In spite of this, the body color pattern observed on our specimens is very close with that reported by Eigenmann (1916)in the original description of  H. landoni. He pointed out that the body has four oblique transverse saddles, the first behind the eyes, the second at the end of the dorsal fin, the third at the level of adipose fin, and the fourth at the caudal-fin origin. On our specimens of  H. landoni, the presence of saddles is variable in that some are very evident while in others they are very faint or no traces are observed ( Fig. 2). Two more species, were originally described from the Pacific versant of Ecuador,  H. hammarlundi Rendahl 1937, and  H. fugleri Ovchynnyk 1971. We agree with they are synonymous with  H. landoniand  H. annectens, respectively. The original description of  H. hammarlundiwas made on a unique specimen of 70.4 mmSL (Armbruster per. comm.). This specimen has small patches of granular plates on belly, and 20 movable odontodes developed in the cheek area. This is clearly different to that observed in  H. furtivus. On the other hand, Ovchynnyk (1971)in the original description of  H. fugleri, points out that the folded dorsal fin reaches or is very close to adipose-fin origin, and that the belly has small patches of granular plates. The specimen used was 91 mmSL. Only the examined juveniles of  H. annectensshow that the folded dorsal fin reaches the adipose-fin origin, and they also have small patches of granular plates on belly. The first character is useful to distinguish juveniles from  H. landoniand  H. furtivus, but the character changes with size; in large specimens of the three species, the folded dorsal fin does not reach the adipose-fin origin. Also, the figures provided by Ovchynnyk (1971)show a color pattern of the body and fins similar to  H. annectens, and different to that observed in  H. furtivus. Additionally, there are no developed cheek odontodes in  H. fugleri. These results support the synonymy proposed by Armbruster et al.(2015) between the species previously described from Ecuador, and confirms the presence of the new species. The completely naked abdomen, the dark rounded dots on it, and the small movable cheek odontodes distinguish  H. furtivusfrom all other species included in  Hemiancistrus sensu lato(Armbruster et al.2015). Finally, Armbruster et al.(2015) established at least three groups within  Hemiancistrus sensu lato. The groupings are based on morphological similarity of species. The species  H. landoniand  H. furtivusare similar in overall body shape and the color pattern, but the cover of the abdomen and the degree development of mobile cheek odontodes are different. Despite these facts both species are included in the  H. landonigroup.   Hemiancistrus annectens,  H. aspidolepis, H. maracaiboensis, H. holostictusand  H. wilsoni, inhabit the trans- Andean region, Panamá, Maracaibo lake basin, the rivers Magdalena, Cauca, Atrato, San Juan (Colombia), and Santiago (Ecuador). When comparing the original descriptions of all species, and specimens of  H. annectens, it is evident that the external morphology, the meristic and morphometric data, and the color pattern are similar. This fact was previously detected by Schultz (1944), and mentioned by Regan (1913)and Eigenmann (1918). Some morphometric differences can be related with the size, the fixation and preservations conditions of the type specimens. At this moment, we do not have specimens of the other species on hand, and we are unable to offer diagnostics characters to recognize each species. A study to establish the status of each species is necessary, using direct comparison of fresh and well preserved specimens from each basin. In the meantime, the use of the names proposed to each basin is suggested. Recently, Armbruster et al.(2015) reassign to the genus  Hypostomusthe five species, as a consequence of the results obtained on phylogenetic analyses. The external morphology of the five trans-Andean species of  Hypostomusseems to indicate that they do not belong in the genus  Hemiancistrusor even to the Ancistrinigroup, as pointed in morphological and molecular phylogenies ( Armbruster 2004, 2008; Lujan et al 2015).   FIGURE 8.External lateral view of right opercular. A)  Hypostomus annectens, specimen of 190.0 mm SL, and B)  Hemiancistrus landoni, specimen of about 210.0 mm SL, without data. The comparative analysis of the bones related to the movement of the opercular, support the placement of  H. annectensin the Hypostomini and  H. landoniand  H. furtivusin the Ancistrini. In order to establish the proper generic allocation of  H. annectensa project is being carried out which includes a comparative analysis of the indicated bones, between this species and species traditionally included in the genus  Hypostomus. In an analogous way, it is hoped to be able to compare osteologically the species included in the genus  Hemiancistrus sensu latoto indicate its appropriate taxonomic status. 1562860351 Ecuador Naranjito 8 229 1 holotype 1562860330 Ecuador Rio Clementina Babahoyo 8 229 1 Los Rios holotype 1562860359 1979-04-13 Material & All & La Cuca & Granja & Barriga Ecuador -3.5 rio Arenillas 21 -80.07222 8 229 1 El Oro Province 1562860357 1979-04-12 Barriga Ecuador -3.636111 rio Arenillas 21 -79.92777 Rio Piedras 8 229 1 El Oro Province 1562860355 2010-11-20 W, P. Tufino & R. Barriga Ecuador -3.648611 Rio Zaracay 8 229 1 El Oro Province 1562860360 1992-09-22 de Noviembre & Barriga Ecuador -2.6905556 Rio Minas 21 -79.63778 8 229 1 Guayas Province 1562860363 1992-09-22 de Noviembre & Barriga Ecuador -2.6905556 Rio Catarama 21 -79.63778 Rio Minas 8 229 1 Los Rios Province 1562860361 1953-05 Estrella Ecuador -1.0344445 Rio Quevedo 21 -79.461945 8 229 1 Los Rios Province 1562860356 1986-06-23 de Jujan & Abril Ecuador Rio Jujan -79.55167 8 229 1 Los Rios Province 1562860358 1983-11 Quevedo, G. Ecuador rio Babahoyo Rio Las Juntas 8 229 1 Los Rios Province 1562860362 1993-02-05 Laguna de Poza Honda & Barriga Ecuador -1.0819445 Rio Portoviejo 21 -80.15445 8 229 1 Manabi Province 1562860364 1992-10-03 Barriga Ecuador -1.1058334 Rio Portoviejo 21 -80.16139 8 229 1 Manabi Province