Revision of the Oxynoemacheilus angorae group with the description of two new species (Teleostei: Nemacheilidae) Author Yoğurtçuoğlu, Baran Hacettepe University, Faculty of Science, Biology Department, Beytepe Campus, 06800 Ankara, Turkey. Author Kaya, Cüneyt Recep Tayyip Erdogan University, Faculty of Fisheries, 53100 Rize, Turkey. Author Freyhof, Jörg 0000-0002-7042-3127 Museum für Naturkunde, Leibniz Institute for Evolution and Biodiversity Science, 10115 Berlin, Germany. joerg. freyhof @ mfn. berlin; https: // orcid. org / 0000 - 0002 - 7042 - 3127 joerg.freyhof@mfn.berlin text Zootaxa 2022 2022-05-06 5133 4 451 485 journal article 55627 10.11646/zootaxa.5133.4.1 c2f66f44-2174-4c3a-82b5-288255ca2cb1 1175-5326 6530889 58A39822-32D1-4907-A866-7E75E5DA70A2 Oxynoemacheilus germencicus ( Erk‘akan, Nalbant & Özeren 2007 ) ( Figs. 7–9 ) Barbatula germencica Erk‘akan, Nalbant & Özeren 2007:70 , Fig. 2 ( type locality: Aydin , Germencik , 15th kilometers, 37°38‘N , 27°18‘E , Turkey ) . Barbatula cinica Erk‘akan, Nalbant & Özeren 2007:73 , Fig. 4 ( type locality: Road of Kütahya to Denizli , Cin Stream , 39°40‘N , 29°30‘E , Turkey ) Oxynoemacheilus mesudae Erk’akan 2012: 98, Figs. 1–2 ( type locality: Büyük Menderes River , Dinar , Civril , Denizli Province , 38°07.366’N , 30°05.723’E , western Turkey ) Material examined. HUIC : BM-3 , paratypes of O. cinicus , 20, 42–58 mm SL; Turkey : Denizli prov. : Cin stream .— FFR 1508 , 22 , 35–65 mm SL; Turkey : Muğla prov. : Çine stream a tributary of Adnan Menderes Reservoir 8 km south of Çine , 37.5427 28.0624 .— FFR 1523 , 7 , 52–58 mm SL; Turkey : Denizli prov. : Aksu stream 4 km north of Honaz , 37.7892 29.2611 .— FFR 1528 , 12 , 39–56 mm SL; Turkey : Denizli prov. : a tributary of Lake Işıklı 1 km north of Çıtak , 38.1558 29.6373 .— FFR 1530 , 61 , 28–68 mm SL; Turkey : Uşak prov. : Banaz River 8 km north of Sivaslı , 38.5501 29.6202 .— FFR 1514 , 2 , 44–47 mm SL; Turkey : Afyonkarahisar prov. : Yeşilhüyük stream 1 km west of Yapağılı , 38.1427 30.0667 .— FSJF 1852 , 21 , 27–54 mm SL; Turkey : Aydın prov. : Yenicay stream in Yenicay , a tributary to River Büyük Menderes , 37.8324 28.5754 .— FSJF 1855 , 10 , 47–64 mm SL; Turkey : Afyon prov. : a spring in Dinar at road from Dinar to Işıklı 38.0705 30.1684 .— FSJF 1875 , 24 , 47–73 mm SL ; FSJF 3025 , 19 , 38–81 mm SL; Turkey : Denizli prov. : spring running to Lake Işıklı at Işıklı , 38.3215 29.8512 .— FSJF 2296 , 9 , 34–55 mm SL; Turkey : Izmir prov. : Çiçekli stream, a tributary to stream Nif at Çiçekli , between Izmir and Manisa , 38.5046 27.2718 .— FSJF 2323 , 13 , 37–55 mm SL; Turkey : Aydın prov. : stream north of Hıdırbeyli dam lake, north of Germencik , 37.9154 27.5984 .— FSJF 2536 , 13 , 56–72 mm SL; Turkey : Afyon prov. : stream west of Kızıllı , northwest of Dinar , 38.1228 30.0954 .— FSJF 2581 , 11 , 36–46 mm SL; Turkey : Aydın prov. : Çine stream south of Eskiçine , 37.5434 28.0630 .— FSJF 3020 , 22 , 46–66 mm SL; Turkey : Denizli prov. : Büyük Menderes at Cindere , 6 km south of Güney , 38.0835 29.0148 .— FSJF 3044 , 7 , 56– 75 mm SL; Turkey : Afyon prov. : Küfü stream near to Saltık , 10 km west of Sandıklı , 38.4838 30.0738 .— FSJF 3056 , 3 , 51–69 mm SL; Turkey : Uşak prov. : Hamam stream at Aksaz , 11 km north of Güney , 38.2982 29.1415 .— FSJF 3081 , 21 , 54–75 mm SL; Turkey : Afyon prov. : Karadirek stream at Başağaç , 10 km north of Sandıklı , 38.5799 30.2018 .— FSJF 3086 , 9 , 62–78 mm SL; Turkey : Afyon prov. : Karadirek stream at Karadirek , 9 km north of Sandıklı , 38.5530 30.1963 .— FSJF 3103 , 25 , 69– 43 mm SL; Turkey : Afyon prov. : Küfü stream 10 km west of Sandıklı , 38.4838 30.0738 .— FSJF 3689 , 13 , 33–61 mm SL; Turkey : Uşak prov. : stream northeast of Banaz , 38.7466 29.7649 .— FSJF 3707 , 22 , 38–62 mm SL; Turkey : Aydın prov. : Akçay River 10 km south of Nazilli , 37.8098 28.3152 . New material used in molecular genetic analysis. FSJF-DNA 2835 ; Denizli prov. : Büyük Menderes at Yenicekent DSI pomps, between Yenicekent and Mahmutlu , 38.0373 28.9636 (GenBank accession numbers: ON123697 , ON123698 , ON123699 ) .— FSJF-DNA 2846 ; Turkey : Uşak prov. : Hamam stream 3 km northeast of Banaz , 38.7466 29.7650 (GenBank accession numbers: ON123700 ) .— FSJF-DNA 2876 ; Turkey : Uşak prov. : Banaz stream at Yenierice , 38.5496 29.6240 (GenBank accession numbers: ON123702 ) . Additional distribution records. Barlas & Dirican 2004 : 37.2135 28.2532, 37.2837 28.1705, 37.3498 28.1024, 37.4531 28.1540, 37.5978 28.0114; Yeğen et al . 2008 : 37.2239 28.8658, 37.2314 37.2314, 37.2356 28.8117, 37.3358 28.7247, 37.4797 28.7300, 37.6333 27.4667, 37.7356 29.5569, 37.9381 28.9867, 38.0800 29.5014, 38.1328 29.4350, 38.1914 29.3875, 38.2331 29.3222, 38.1200 29.0461, 38.5194 30.1753, 38.6247 30.2903; Van Neer et al . 2008 : 37.7170 27.5540, 37.9858 28.9750, 38.2052 30.0468, 38.2422 29.1782, 38.3257 29.8075; Geiger et al . 2014 : 38.1300 30.2000 ; Erk’akan 2012: 38.1306 30.1956 ; Güçlü et al . 2013 : 37.4403 28.3806, 37.4713 29.1153, 37.5553 27.4641, 37.6517 27.9975, 37.7075 27.4761, 38.0219 28.5746, 38.0938 29.1223, 38.2561 29.9222, 38.2713 29.5931; Yoğurtçuoğlu et al . 2021a : 37.4261 28.1376 ;; unpublished records: 37.7832 27.8390, 37.9100 27.6000, 38.1268 30.0854. Diagnosis. Oxynoemacheilus germencicus is distinguished from other species of the O. angorae group by a combination of characters, none of them unique. It is distinguished from O. anatolicus , O. angorae and O. eregliensis by having small embedded scales on the belly, often restricted to an area between the pelvic-fin bases (vs. belly without scales); an emarginate caudal fin in which the middle caudal-fin ray is 72–82% of the length of the longest caudal-fin ray (vs. slightly emarginate or almost truncate, 84–93 in O. anatolicus , 88–92 in O. angorae , 83–91 in O. eregliensis ). Oxynoemacheilus germencicus is further distinguished from O. anatolicus by having a shallower head (head depth at eye 41–49% HL, mean 45% vs. 47–57, mean 50), and smaller interorbital distance (27–34% HL vs. 33–40), and from O. angorae by the absence of midlateral stipe (vs. usually present), no depigmented stripe along the anterior part of the lateral line (vs. present), and by a greater pre-pelvic length (51–57% SL vs. 48–51). It is further distinguished from O. eregliensis by having a longer anal fin (anal fin length 16–19% SL vs. 14–16), and a shallower head (head depth at eye 41–49% HL vs. 48–57). Distribution. Oxynoemacheilus germencicus is widespread in the Büyük Menderes River drainage. It is also recorded from one spot in the Gediz River drainage (Nif subdrainage, FSJF 2296), a record that needs to be confirmed in the future. Remarks. Oxynoemacheilus germencicus is highly variable in its colour pattern ( Fig. 7 ) ranging from a marbled pattern with large roundish blotches, to a fine mottled pattern, some individuals having or lacking bars on the flank. The variability of the colour pattern is stronger between populations, individuals within populations being more similar to each other. There are also some morphometric differences among the populations, especially in the shape of the caudal fin, and the depth of the body and caudal-peduncle (see below). Here again, the variability between populations is stronger than that within populations. The intraspecific variability (maximum barcode distance) within O. germencicus is 1.2%. When compared to 2.2% COI distance in O. angorae , O. germencicus has the second-highest intraspecific genetic variability of the members of the O. angorae group. Freyhof et al . (2022) gave the intraspecific range of the K2P distance of DNA barcode data from 53 Oxynoemacheilus species ranging from 0.0–2.4% and species of the O. angorae group range within these boundaries. The morphological and genetic variability might be indicative of locally isolated and adapted populations in O. germencicus . We do not find the morphological variability to be connected to subsystems within the Büyük Menderes drainage but to macrohabitats, where slender individuals with a more deeply emarginate caudal fin are found in small streams with high winter floods (as seen from stream shape) and individuals with deeper body are found more in springs and other slowflowing waterbodies. It would be worthwhile to study this case more in detail in the future. Indeed, all morphometric and colour pattern characters are largely overlapping. While we find populations with non-overlapping character states, other populations are intermediate, bridging the small gaps. Erk’akan et al . (2007) gave the type locality of O. germencicus at 37°38’N 27°28’E , which is 20 km southwest of Germencik. There are no streams at the points indicated by the coordinates. We found the species north of Germencik (FSJF 2323), which is given as the type locality. We suspect that the coordinates published by Erk‘akan et al . (2007) are a lapsus. See also Kottelat (2012) for discussion about coordinates and type localities of other species described by Erk’akan et al . (2007) . Erk’akan et al . (2007) provided the type locality of O. cinicus at 39°40’N , 29°30’E , a place in Simav River drainage, where the species was never found. It was corrected as 37°40’41.07’’N 29°30’13.08’’E by Erk’akan (2012). The type locality is situated at a small stream flowing to the upper Emir, a major tributary of the Büyük Menderes. Müfit Özuluğ ( İstanbul ) and JF visited the location in 2008 but found no water in the stream. Erk’akan et al. (2007:73) distinguished O. cinicus from O. germencicus by: “absence of a sexual dimorphism and suborbital flap, maximum depth of the body and caudal peduncle; caudal fork is less forked and colour pattern”. Eighteen of the 23 paratypes ( 40–57 mm SL) of O. cinicus (HUIC: BM-3) had already been dissected and their sex was determined by their gonad structure (BY, own examination). We found only one male and 17 females among these paratypes . The other five individuals were not dissected. In the paratypes of O. cinicus (HUIC: BM-3), the pectoral fin in females is 20–23 % in SL (n=8) and 26% in the single male identified (lowermost individual in Fig. 8 ). Sexual dimorphism in the length of paired fins is one of the diagnostic characters in the genus Oxynoemacheilus and is found in all species of the O. angorae group including O. cinicus . Furthermore, we found a narrow suborbital groove in the male, clearly demonstrating that there is sexual dimorphism and a suborbital groove in O. cinicus . FIGURE 7. Oxynoemacheilus germencicus , left column, from top: FSJF 2323, 54.2 mm SL; stream north of Germencik; not preserved, about 60 mm SL; same place as FSJF 3056, Hamam stream; FSJF-DNA 2846, about 55 mm SL; Hamam stream; FSJF-DNA 2827, about 60 mm SL; Büyük Menderes River at Yenicekent; right column, from top: FSJF 3025, 79 mm SL; Işıklı spring; FSJF 3081, 70 mm SL; Karadirek stream; not preserved, about 50 mm SL; same place as FSJF 3020, Büyük Menderes at Cindere; FSJF 2581, 46 mm SL; middle Çine stream. Erk’akan et al . (2007:73) gives the range of the caudal peduncle depth as 10.6–12.5% SL in O. cinicus (vs. 9.2–11.5 in O. germencicus ). We measured the caudal peduncle depth as 9.2–10.9% in FSJF 2536 (n=13), 10.3– 12.1% SL (FSJF 2323, n=7), 10.4–12.2% SL (FFR 1508, n=7), 11.8–12.7% SL (FFR 1530, N=9), and 10.0–11.8% SL (FFR 1523, n=7) in O. germencicus . The caudal peduncle depth already overlaps between O. cinicus and O. germencicus in the data by Erk’akan et al . (2007:73) and our measurements almost exactly agree with the data given by Erk’akan et al . (2007:73) for O. cinicus . Erk’akan et al . (2007:73) gives the range of the body depth as 11.8–13.3% SL in O. cinicus (vs. 15.7–17.0% SL in O. germencicus ). However, our measurement of the 20 type specimens (HUIC BM-3) resulted in 13.0–16.4% SL, clearly placing them in the range of O. germencicus and not in O. cinicus . In our materials of O. germencicus , we measured a body depth of 12.8–15.6% SL (FSJF 3103, n=12), 15.0–17.8% SL (FSJF 2323, n=7), 14.5–19.4% SL (FSJF 2536, n=13), 15.6–18.3% SL (FFR 1508, n=7), 15.2–17.5% SL (FFR 1523, n=7), and 15.6–17.6% SL (FFR 1530, n=8). The body depth of O. germencicus from Germencik (FSJF 2323) agree with the data published for this species by Erk’akan et al . (2007:73) . Other populations are more or less slender: body depth is a character that is variable between populations, potentially depending on the local environmental conditions, gonadal and conditional states, and other adaptations. We do not consider the body depth to be a suitable character to distinguish O. germencicus and O. cinicus . Taken together, the caudal peduncle depth in O. cinicus and O. germencicus overlaps substantially and cannot be used as a diagnostic character. FIGURE 8. Oxynoemacheilus germencicus , HUIC : BM-3, paratypes of O. cinicus , from top: 57 mm SL; 49 mm SL; 46 mm SL; 44 mm SL; Cin stream. In addition, we found no difference in the caudal fin furcation and no differences in the colour pattern, which the latter is so variable in O. germencicus . We found also no other character to distinguish the examined paratypes of O. cinicus (HUIC: BM-3) from O. germencicus and therefore we treat both species as conspecific. Oxynoemacheilus germencicus and O. cinicus were described in the same publication, and as First Revisers, we give priority to O. germencicus over O. cinicus and treat O. cinicus as a synonym of O. germencicus . Erk’akan (2012) described Oxynoemacheilus mesudae from Dinar, in the upper Büyük Menderes drainage, the type locality is about 5 km north of Dinar at 38°07.36’N 30°05.72’E . Erk’akan (2012:99) distinguished O. mesudae from O. germencicus by: “its shape of bony swim-bladder capsule, digestive track [sic] and some morphometric data”. We examined the shape of the bony swim-bladder capsule in five individuals each of O. germencicus (FSJF 2323) and O. mesudae (FSJF 2536) ( Fig. 9 ). Although there is some variation in the shape of the capsule within the populations, no consistent inter-population difference could be detected. The digestive tract in O. germencicus is not shown or described by Erk’akan et al . (2007) or Erk’akan (2012). It remains unclear how the species might be distinguished by this character. We examined the digestive tract in two individuals each from the type locality of O. mesudae (FSJF 2536) and the type locality of O. germencicus (FSJF 2323) but found no difference. We think it unlikely that differences will be found even if more individuals are examined. Additionally, we found no differences in colour pattern between the two species. Erk‘akan et al . (2007) and Erk’akan (2012) examined only eight individuals of O. germencicus . They have obviously strongly underestimated intraspecific variability in the colour pattern of this species. While the differences in morphometric characters are not specified by Erk’akan (2012), we extracted these data from the description of O. germencicus by Erk’akan (2007:71) and Table 1 given by Erk’akan (2012:98) in the description of O. mesudae . Based on these data, O. mesudae is distinguished from O. germencicus by the length of dorsal-fin base, predorsal length, pre-anal length, caudal peduncle length, eye diameter, and interorbital distance. We measured these characters in our materials from the type localities of O. mesudae (FSJF 2536 and FFR 1528) and O. germencicus (FSJF 2323) ( Table 2 ). Our measurements often disagree with the data given by Erk‘akan (2007:71) and Erk‘akan (2012:98). This could be related to differences in measuring methodologies. We find no consistent characters differentiating populations described as O. mesudae from O. germencicus in our own measurements. Given that both are not distinguished even by their colour pattern and COI sequence data, we treat O. mesudae as a synonym of O. germencicus .