Zoosyst. Evol. 97 (1) 2021, 181-189 | DOI! 10.3897/zse.97.62120 gee BERLIN The discovery of a microbialite-associated freshwater fish in the world’s largest saline soda lake, Lake Van (Turkey) Mustafa Akkus!, Mustafa Sari*, F. Giiler Ekmekci?, Baran YoSurtcuoglu? Yiiziincii Yil University, Faculty of Fisheries Science, Bardak¢1, Tusba, Van, 65090, Turkey 2 Bandirma Onyedi Eyliil University, Maritime Faculty, Sehit Astsubay Mustafa Soner Varlik Avenue No:77, Bandirma, Balikesir, 10200, Turkey 3 Hacettepe University, Faculty of Science, Biology Department, Beytepe Campus, Cankaya, Ankara, 06800, Turkey http://zoobank. org/7127F 55 D-ED8F-4CF6-A537-1SFB4F91A765 Corresponding author: Baran Yogurtcuoglu (yokbaran@gmail.com) Academic editor: Nicolas Hubert Received 14 January 2021 # Accepted 1 March 2021 # Published 16 March 2021 Abstract Lake Van is the largest saline soda lake in the world and one of the world’s few endorheic lakes of greater than 3,000 km surface area. Despite its huge size, no fish species have so far been known to permanently occur in this lake due to its extreme environmental conditions. Here, we report the discovery of a fish population that permanently inhabits some of the unique microbialites of the lake, at a maximum depth of 13 m and about 500 m offshore. We tested whether this is an undescribed species or a new occurrence of a known species. A molecular and morphological examination showed that the newly discovered fish represents an isolated population of Oxynoemacheilus ercisianus, the only nemacheilid loach native to the freshwater tributaries of the Lake Van endorheic basin. Our further hypotheses on the prediction that (a) stream fishes would have a more anterior placement of fins than lake fishes were supported; but, that (b) stream fishes would be more slender bodied than their lake conspecifics was not supported. The lake dwelling population also shows very small sequence divergence (0.5% K2P distance) to its stream dwelling conspecific in the mtDNA-COI barcode region. The notable morphological difference with minute molecular divergence implies that the newly discovered popu- lation might have lost its link to freshwater during desiccation and transgressional phases of the Lake Van, and has adapted to a life on the microbialites. Key Words Biodiversity, COI, Eastern Anatolia, extreme environments, morphology, Nemacheilidae, Oxynoemacheilus ercisianus Introduction Soda lakes are characterized as extreme environments since they are highly alkaline, originate in closed basins, and are often exposed to high evaporation rate (Jones et al. 1998; Pinti 2014). This evaporation rate results in elevated concentration of dissolved salts, especially of [CO,”] and, in turn, increased salinity and pH levels that usually reach between 8.5 and 11.5 (Pinti 2014). Soda lakes occur all over the world with the best-known examples found along the East African rift and western USA (Schagerl and Burian 2016). However, Lake Van, which is located in the uplands of eastern Anatolia, is a less known soda lake that is famous for being the largest soda lake, and the third largest closed lake on Earth (Kempe and Kazmierczak 2011). Lake Van is also unique because it possesses the largest recent organosedimentary deposits (microbialites) on earth (Kempe et al. 1991). The lake is located over 1640 m above sea level and has a maximum depth of 450 m, 9.7-9.9 pH and 22%bo total salinity (Degens et al. 1984; Reimer et al. 2009). Due to these extreme living conditions, none of the fish species in the endorheic lake basin 1s expected to enter the lake itself. Indeed, all the fish species recorded in the basin, namely Alburnus timarensis, Barbus lacerta, Capoeta damascina, Oxynoemacheilus ercisianus and non-native Salmo trutta, Copyright Mustafa Akkus et al. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which per- mits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. 2 permanently inhabit the freshwater tributaries and streams outside the lake (Elp et al. 2016; Kaya 2020). One exception to this is the diadromous population of A/burnus tarichi (Tarek), which is able to enter the estuaries and the lake, but still migrates to freshwater tributaries on which it relies to spawn in the reproduction season (Sari 2008). Our knowledge in this regard has become enhanced after an exploration in a microbialite area in the south of the lake by scuba-diving operations approximately 500 m offshore. Here, we surprisingly discovered a nemacheilid fish occupying the branches and holes of a 13-meter high tower-like microbialite in 15-meters depth of the lake. The fish seemed morphologically different from Oxynoemacheilus ercisianus, the only nemacheilid loach species in the Lake Van basin. Therefore, our main hypothesis was to test if the newly-found lake population of nemacheilid fish is conspecific with O. ercisianus, or it might represent an undescribed species. This was addressed by producing mtDNA-COI barcode sequences and by examining some morphological traits. Nemacheilid loaches are typically found in shallow waters and associated with fast to moderate flowing stretches of the streams, and to a lesser extent, large rivers and lake shores. Habitat associated divergence in mor- phological traits has been documented in a great number of studies with the best known examples including spe- cies from Characidae, Cichlidae (Langerhans et al. 2003; Costa-Pereira et al. 2016; Perazzo et al. 2019); Cyprini- dae (Haas et al. 2010; Franssen et al. 2013); Galaxiidae (Dunn et al. 2020) and Centrarchidae (Brinsmead and Fox 2002). To the best of our knowledge, no study has yet addressed this topic in nemacheilids, and this study is also the first one to document morphological differences between two groups to see if the differences are due to a response to the contrasting habitats (lentic vs. lotic), or imply species-specific characters. Material and methods Fish sampling and measurements Fish were collected by scuba diving using a small dip net with 4 mm mesh size, by two sampling events carried out in January 2018 and January 2020. After anaesthe- sia, fish were fixed in 5% formaldehyde and stored in 70% ethanol, or directly stored in 99% ethanol. Measure- ments were made with a digital calliper and recorded to 0.1 mm. All measurements were made point-to-point, never by projections. Methods for counts and measure- ments follow Kottelat and Freyhof (2007). Standard length (SL) is measured from the tip of the snout to the posterior extremity of the hypural complex. The length of the caudal peduncle is measured from behind the base of the last anal-fin ray to the posterior extremity of the hypural complex, at mid-height of the caudal-fin base. The last two branched rays articulating on a single pte- rygiophore in the dorsal and anal fins are counted as 1%. zse.pensoft.net Akkus, M. et al.: Microbialite-associated freshwater fish in Lake Van Simple rays of dorsal and anal fins are not counted as they are deeply embedded. Our hypothesis does not include testing for the varia- tions in body shape. Therefore we used only linear meas- urements to identify morphological differences between the two groups of fish collected from stream and lake. All body measurements were standardized by individu- als’ SL, and the measurements taken from the head re- gion were standardized by individuals’ head length. To reduce the effect of allometric growth on some ratios, we selected similar size range for comparison (i.e., 30.0 mm to 39.2 mm for lake and 30.7 mm to 42.4 mm for stream groups). We also examined univariate patterns to test for significant effect of the explanatory variable as habitat type (stream vs. lake) on each of the response variables (morphometric measurement) using regression model with analysis of covariance (ANCOVA). In this proce- dure, SL was used as a covariate to control for variation due to fish size (all dependent variables and the covariate were log-transformed) (Zar 2010). Statistical effects eval- uated at a = 0.01. The water physicochemical parameters including tem- perature, pH, dissolved oxygen and salinity were meas- ured at the sampling locations (both in streams and in the lake) using a multiparameter instrument (YSI ProDS, Yellow Spring Instruments, Yellow Springs, OH, USA). The lake measurements were taken during scuba diving, just near the surface of the microbialites where the fish were collected, and not from the open water. Abbreviations used: SL, standard length; HL, head length. Collection codes: FFR, Recep Tayyip Erdogan University Zoology Museum of the Faculty of Fisheries, Rize, Turkey. Molecular data analysis Genomic DNA was extracted from fin tissue using Mach- erey and Nagel NucleoSpin Tissue kits following the manufacturer’s protocol on an Eppendorf EpMotion® pi- petting-roboter with vacuum manifold. The standard ver- tebrate DNA barcode region of COI was amplified using a M13-tailed primer cocktail including FishF2_tl (5’TG- TAAAACGACGGCCAGTCGACTAATCATAAAGA- TATCGGCAC), FishR2 tl S’CAGGAAA- CAGC- TATGACACTTCAGGGTGACCGAAGAATCAGAA), VF2 tl (5’TGTAAAACGACGGCCAGTCAAC- CAACCACAAAGACATTGGCAC) and FRId _tl (S’;CAGGAAACAGCTATGACACCTCAGGGTGTCC- GAARAAYCARAA) (Ivanova et al. 2007). PCR were performed using Qiagen Multiplex taq as follows: 15min at 95 °C; 10 cycles of 35 s at 94 °C, 90 s at 52-49 °C (touch-down) and 90s at 72 °C followed by 25 cycles of 35 s at 94 °C, 90s at 55 °C and 90s at 72 °C with final elongation for 10min at 72 °C and hold at 10 °C. Sequencing of the ExoSAP-IT (USB) purified PCR prod- uct in both directions was conducted at Macrogen Eu- rope Laboratories with forward sequencing primer M13F Zoosyst. Evol. 97 (1) 2021, 181-189 Oxynoemacheilus ercisianus MW684713 Oxynoemacheilus ercisianus MW684715 84-78 —— 0.0010 Oxynoemacheilus ercistanus KU928283 3 Oxynoemacheilus ercisianus MK546453 80-79] Oxynoemacheilus ercisianus MK546452 Oxynoemacheilus ercistanus MK546454 Stream Oxynoemacheilus ercisianus MH469268 Oxynoemacheilus ercisianus MK546485 Oxynoemacheilus ercisianus MH469267 62-65! Oxynoemachetlus ercisianus MK546484 Oxynoemacheilus ercistanus MW684714 Lake Oxynoemacheilus ercisianus MK546487 Oxynoemacheilus ercisianus MK546488 Stream Oxynoemacheilus ercistanus MK546486 Figure 1. Maximum likelihood estimation of the phylogenetic relationships based on the mitochondrial COI barcode region (K2P model, discrete Gamma distribution for rate differences with 3 categories (+G, parameter = 0.0500)). Nucleotide positions with less than 98% site coverage were eliminated, resulting in 570 analysed positions. Numbers of major nodes indicate bootstrap values from 1000 pseudoreplicates from the NJ and ML method. (S’;GTAAAACGACGGCCAGT) and reverse sequencing primer M13R-pUC (S°CAGGAAACAGCTATGAC). Molecular analysis involved 16 nucleotide sequences. Among these, we newly generated 3 DNA barcodes from lake dwelling specimens and included already published data from NCBI GenBank for 11 specimens from Ilica (Zilan) stream, a northern tributary in Ercis (Geiger et al. 2014) (Table 1). The Muscle algorithm (Edgar 2004) was used to align the DNA barcodes after manually screening for unexpected indels or stop codons. The sequence evo- lution model test implemented in the MEGA X software (Kumar et al. 2018) was used to determine the most ap- propriate evolution model for the given data and to recon- struct the mitochondrial relationships between the studied groups. The model with the lowest BIC scores (Bayesian Information Criterion) is considered to best describe the substitution pattern. Initial tree(s) for the heuristic search was obtained by applying the Neighbour-Joining meth- od to a matrix of pairwise distances estimated using the Maximum Composite Likelihood approach. A discrete Gamma distribution was used to model evolutionary rate differences among sites (5 categories (+G, parameter = 0,0500)). We generated neighbour-joining (Saitou and Nei 1987) and maximum likelihood (ML) phylogenetic trees with 1000 bootstrap replicates to explore phyloge- netic affinities of the mitochondrial lineages. The tree is drawn to scale, with branch lengths measured in the Table 1. List of COI sequences downloaded from NCBI Gen- Bank with information on drainage and country of origin. Species Drainage Country Genbank Reference Oxynoemacheilus llica Turkey MH469267 Turan et al. 2019 ercisianus Stream MH469268 (Van) MK546484 Geiger MF MK546485 (unpublished) MK546486 MK546487 MK546488 Magara MK546452 Stream MK546453 (Van) Mk546454 KU928283 Freyhof et al. 2016 number of substitutions per site. All codon positions were included and positions with less than 98% site coverage were eliminated, resulting in 570 analysed nucleotide po- sitions. Evolutionary analyses were conducted in MEGA X (Kumar et al. 2018). Results In total, 44 fish from microbialites and 33 fish from streams were collected and examined. The standard length (SL) of the Oxynoemacheilus individuals from microbialites ranged from 17.8 mm to 39.2 mm, whereas the size of the stream group ranged from 30.7 mm to 66.7 mm. zse.pensoft.net Akkus, M. et al.: Microbialite-associated freshwater fish in Lake Van Figure 2. A lake-resident individual of Oxynoemacheilus ercisianus on the microbialite (left). Views from microbialites in Lake Van (right). Molecular and morphological assessments The analysis of the nucleotide sequences of the COI bar- code region resulted in a mean 0.5% K2P distance be- tween the stream and the lake-resident groups, both are separated by 7 variable nucleotide substitutions one of which is unique to the lake group. We treated, therefore, the lake-resident group as conspecific to O. ercisianus. Despite the low molecular distance, several signifi- cant differences were found between the morphological characters of the two populations. The lake-resident pop- ulation is differentiated from the stream population by the following combination of morphological characters: longer pre-dorsal length (52-56% SL vs. 50-52), longer pre-anal length (78-83% SL vs. 73-77), longer pre-pel- vic length (58-61 %SL vs. 52-55) and greater distance between pectoral and pelvic fin origin (33-36% SL vs. 26-31), shorter pelvic fin length (11-13% SL vs. 14-16) and shorter caudal peduncle length (14-16% SL vs 16— 18). The pectoral fin length is significantly greater in the stream dwelling population, yet the range overlapped with lake-resident population. Similarly, the snout, post-orbital and inter-orbital distances were significantly greater in the stream population, whereas all overlapped by means of minimum and maximum ratio. See Figs 2, 3 and Table 2 for general appearance and comparison with stream population and other morphometric measure- ments, respectively. The lake-resident population is further differentiated from the stream population by having a shorter lateral zse.pensoft.net Table 2. Nucleotide substitutions in the variable sites of the mi- tochondrial COI gene (570 bp) of Oxynoemacheilus ercisianus from lake and stream populations. Nucleotide position 1344 343888 Individuals S Shove O227.5 Oxynoemacheilus ercisianus MW684715 (Lake) TCCCTGA Oxynoemacheilus ercisianus MW684714 (Lake) ee ioe: Oxynoemacheilus ercisianus MW684713 (Lake) oe Oxynoemacheilus ercisianus KU928283 (Stream) CC... ..G Oxynoemacheilus ercisianus MK546487 (Stream) - AA. . . Oxynoemacheilus ercisianus MH469267 (Stream) S Repco Ge. Oxynoemacheilus ercisianus MH469268 (Stream) ae ce Oxynoemacheilus ercisianus MK546488 (Stream) Brags Oxynoemacheilus ercisianus MK546486 (Stream) - AA. .A. Oxynoemacheilus ercisianus MK546485 (Stream) Sees omg Gs Oxynoemacheilus ercisianus MK546484 (Stream) - . A.C .G Oxynoemacheilus ercisianus MK546454 (Stream) C . _G Oxynoemacheilus ercisianus MK546453 (Stream) C . x NG Oxynoemacheilus ercisianus MK546452 (Stream) C . rec) line with 5—7 pores reaching up to vertical of pectoral fin midline (vs. 9-12 pores reaching up to pectoral fin tip or dorsal fin origin). Usually, there are none, or just one faint lateral pore in supratemporal canal (vs. two or three apparent pores). All other meristic traits includ- ing fin ray numbers overlapped between lake and stream dwelling populations. Zoosyst. Evol. 97 (1) 2021, 181-189 Figure 3. Oxynoemacheilus ercisianus, left from top: FFR 01403, 32.3 mm SL, 31.1 mm SL, 36.8 mm SL, 37.3 mm SL, Edremit (La- ke-resident population); right from top: 37.4 mm SL, 42.0 mm SL, 41.0 mm SL, 33.4 mm SL, Bendimahi River (Stream population) The SL of the lake-resident population ranged from 17.8 mm to 39.2 mm, suggesting smaller maximum size compared to the stream population (ranged from 30.7 mm to 66.7 mm SL). Material examined Oxynoemacheilus ercisianus (stream dwelling popula- tion): FFR 15533, 3, 35-48 mm SL, Turkey: Van prov.: II- ica stream at Ercis, 2 km northwest to Ulupamir, 39.1813, 43.3019. -FFR 15534, 3, 38-49 mm SL; Turkey: Van prov.: Ilica stream at Ercis Orene, under the bridge at Bit- lis-Tatvan Road, 39.0063, 43.3180. —Uncat., 27, 31-67 mm SL, Turkey: Van prov.: Ilica stream 20 km north of Ercis, 39.2264, 43.3887. Oxynoemacheilus ercisianus (lake-resident popula- tion)—Uncat., 19, 18-36 mm SL, Turkey: Van prov.: gulf near Gevas in the southern Lake Van, 38.3168, 43.1149. —FFR 01403, 25, 19-39 mm SL, Turkey: Van prov.: shore of Edremit, Lake Van, 38.4250, 43.2319. Material used in molecular genetic analysis Uncat. Turkey: Van prov.: gulf near Gevas in the southern Lake Van, 38.3168, 43.1149 (GenBank accession num- bers: MW684713, MW684714, MW684715). Habitat characteristics The lake-resident populations of O. ercisianus were found from two microbialite areas on the south-eastern coast of the Lake Van (Fig. 4), from approximately 0.5 km offshore to the west of Edremit and from 0.3 km off- shore to the north of Gevas in the south-eastern Anatolia. The specimens were collected from a 13-meter-high tow- er-like microbialite under 15-m depth from the surface in Edremit and from a 3-meter-high tower-like microbialite under 6-m depth in Gevas. The microbialites are covered by several groups of algae, and they have several holes, cavities and indented portions where fish are found. The water coming out of the cracks in the lake bottom fol- lows through the microbialites and leaks at the top like a spring water coming out of the crevasse. Leaked water has formed a microhabitat on the microbialite surface and the water physico-chemical parameters fluctuated in par- allel with the amount of this leakage. However, we were able to measure some of the water parameters during scuba-diving for sampling (Table 3). During the daytime dives, very few fish were able to be seen on the micro- bialite, while hundreds of fish were seen in the night dives. Based on this observation, it might be concluded that the fish exhibits a nocturnal activity. All the fish were observed just on the near surface branches of the micro- bialites; and they were restricted to maximum ca. 1 m diameter radius of movement range. No individual was observed at the bottom surface of the microbialites or in zse.pensoft.net 6 Akkus, M. et al.: Microbialite-associated freshwater fish in Lake Van N + Bendimahi River Tlica/River Ercis * 2 Oxynoemacheilus ercisianus (Lake) tr Oxynoemacheilus ercisianus (Stream) @ City Centers Edremit —y Streams & rivers @®p Lakes @ Gevas Si Figure 4. Map of the study area and sampling locations in the Lake Van basin. VAN Tatvan Table 3. Morphometric data of Oxynoemacheilus ercisianus from lake and stream populations (Lake, FFR 01403, n = 12; Stream, FFR 15533, n= 10). Bold text — both significant at p < 0.01 (ANCOVA) and non-overlapping ranges. Lake-resident population Stream population Significance mean min max SD mean min max SD p-Value Standard length (mm) 30 39 cal 42 In percent of standard length Head length 24.7 Zo 26.4 1.0 25.7 24.0 27.6 1.0 0.020 Body depth at dorsal-fin origin 15.9 14.0 17.3 183 MA? 16.3 20.1 de 0.266 Pre-dorsal length 53.8 52.1 56.0 te 50.8 49.8 51.7 0.8 < 0.001 Pre-pelvic length 59.2 58.2 60.7 0.8 53.2 51.9 54.7 1.0 < 0.001 Pre-pectoral length 28.9 28.4 29.6 0.4 28.3 27.1 29:5 0.9 0.233 Pre-anal length 80.0 77.8 82.8 A 74.9 73.1 77.0 1.2 < 0.001 Post-dorsal length Lome h Baa 349 1.0 S70 35:0 38.6 1.3 0.003 Distance between pec. and pel-fin origins 34.6 33.0 35.7 1.0 29.2 26.4 30.6 1:3 < 0.001 Distance between vent and anal-fin origin 29 age Re ge: 0.3 S2 2.8 ae 0.3 0.490 Distance between pel. and anal-fin origins a hie LZ 2109 0.8 21.4 20.1 22.1 nee 0.638 Dorsal-fin length eS 16.2 195 1.0 20.2 LOA 2157 0.9 < 0.001 Anal-fin base length wae: 6.7 9.0 0.6 8.2 6.9 9.0 0.6 0.012 Pectoral-fin length 17.6 16.0 19.7 3 20.2 18.2 22.0 1.4 0.006 Pelvic-fin length 12.2 11.2 13.3 0.7 14.8 13.6 15:9 0.8 < 0.001 Length of caudal peduncle 14.5 13.5 16.1 0.8 17.0 16.2 18.2 0.6 < 0.001 Depth of caudal peduncle Gal 8.2 10.1 0.7 10.2 9.6 10.8 0.5 0.012 In percent of head length Snout length 35.9 34.4 32.5 1.0 38.0 36.3 40.0 1.6 0.003 Eye diameter 22 al 19.9 25.0 13 20.4 18.6 22.6 1.4 0.565 Interorbital width 33:2 30.6 Soul 1.4 8548 S23/ Sr15 1.8 < 0.001 Postorbital distance 45.0 42.3 47.4 1.7 47.2 45.7 48.6 1.1 0.001 Maximum head width 63.4 61.8 65.4 se! 64.6 61.8 70.2 2.8 0.065 Head depth at eye 45.9 43.8 48.0 le? 49.0 45.2 oy 2.4 0.005 Length of inner rostral barbel 17.1 1653 13 1.0 16.4 14.4 18.2 ics 0.832 Length of outer rostral barbel Ly 17.3 21.2 1.3 AS 16.3 S56 1.2 0.363 Length of maxillary barbel 2272 20.7 23.6 1.0 20.1 18.4 22,7 1.4 0.070 zse.pensoft.net Zoosyst. Evol. 97 (1) 2021, 181-189 Table 4. Water physico-chemical parameters measured in two microbialite areas (in Edremit and Gevas and in Bendimahi river. Parameter Edremit Gevas Bendimahi (Lake) (Lake) (Stream) Water temperature (C°) TL 12.3 6.8 Salinity (%o) 18.1 18.3 0.4 pH O51 9.2 AQ Dissolved Oxygen (mg/L) 746 ED 10.2 Total Dissolved Solids (g/L) ONS 20.1 1536 open water. More explorations are needed to demonstrate whether the species is restricted to two microbialite areas in the lake or it is more common on this type of ground- water-fed microbialites. Discussion The findings of the present study have entirely changed the generally accepted knowledge that no fish species is found to permanently occur in Lake Van. Our discovery of the first fish, a nemacheilid loach, permanently inhab- iting the lake, triggered the question whether it might be a new or undescribed species. Our hypothesis testing resulted in recognizing the fish as a distinct and isolat- ed population of Oxynoemacheilus ercisianus, the only nemacheilid species in the endorheic Lake Van basin. Despite considerable differences in some of the morpho- metric characters between the newly found lake-resident population and its stream conspecific, the two groups are superficially very similar to each other with also very small K2P distance in their COI barcode region. We fol- low Freyhof et al. (2018) and Yogurtcuoglu et al. (2020) treating populations without clear morphological differ- ences as conspecific, if they have K2P distances smaller than 2% separating them. However, if diagnostic differ- ences are observed, then small K2P distance is violat- ed and the entities are treated as separate species. In the present study, the morphological differences between the two groups indicated a response to the contrasting habi- tats (lentic vs. lotic), rather than being species-diagnos- tic difference. Indeed, several studies have demonstrated that populations of the same species inhabiting different flow regimes (e.g. lentic vs. lotic) may have a different set of morphological characters. For example, it has been hypothesised that the lateral fins of lake populations will be more posteriorly positioned than those of stream pop- ulations of a fish species (Webb 1984; Swain and Holtby 1989; Langerhans 2008). This prediction 1s supported by the pelvic fins of the lake-resident population being re- markably more posteriorly positioned than in the stream population. This was not supported for the pectoral fins, and it resulted in greater distance between pectoral and pelvic fins in lake-resident population. McGuian et al. (2003) demonstrated that some lake-dwelling rainbow fish species (Melanotaenia spp.) had significantly more posteriorly positioned dorsal fins relative to their stream dwelling conspecifics. This was also supported in O. er- cisianus, as the dorsal fin of the lake-resident population was significantly more posteriorly positioned relative to stream dwelling population. According to Webb (1984) the more posterior position of the lateral fins allows for additional manoeuvrability in fishes. Yet, McGuian et al. (2003) accepted that their findings (more posterior dorsal fin in lake populations) were inconsistent with the expec- tations of hydromechanical evolution, and they avoided proposing a hypothesis of divergence in this trait of fin positioning, as 1t might have been driven by complicat- ed factors such as predator occurrence or genetics. We partly excluded this explanation as the lake is free of any predators, and also our genetic analysis is not able to sup- port or oppose this prediction. According to the hydrody- namic theory, stream fishes would be more slender-bod- ied than their lake conspecifics to minimize drag induced by the water current (Webb 1984; McLaughlin and Grant 1994). This prediction was not supported by our dataset, as we found no significant difference in body depth be- tween two compared populations. A possible explanation for this might be that this prediction has been generally tested and associated with fishes that maintain sustained swimming, like salmonids; whereas Oxynoemacheilus species generally lack sustained swimming ability. To answer how O. ercisianus might have been locked in the microbialites of the Lake Van is not easy. Howev- er, according to the geological support, the Lake Van had been exposed to a combination of rapid desiccation and transgressional phases throughout the Holocene and late Pleistocene (Reimer et al. 2009), and we propose that the newly found lake-resident population might have lost its link to freshwater during these historical water level fluc- tuations. These rapid changes in the lake’s level have also been demonstrated to lead to the development of relic del- tas 40-60 km away from the present river mouths (Degens et al. 1984), which overlap the limits of the microbialite range. As a result, one possible scenario explaining the confinement of O. ercisianus in the microbialites is that, during these lake level fluctuations in the past, some fresh- water populations might have been stuck in one or more of the coastal aquifers that were further inundated by the lake, on which the microbialites were developed. As a pri- mary freshwater group, nemacheilids are intolerant to high range of water salinities. Therefore, the lack of connection between the microbialites to the freshwater streams is hin- dering fish to migrate through the extreme ionic gradient. Indeed, we observed individuals in only a very restrict- ed range of motion where they rely upon the freshwater seepage on the microbialites to survive. They cannot even move to the base of the microbialites, where no or little freshwater comes out. The occurrence of juveniles caught in January (probably the young of the year) also supports the case that the fish has become well-adapted to a micro- bialite-associated life. Its nocturnal behaviour, small size and a very restricted occurrence also might have impeded its discovery by the earlier explorations. Further research is needed to better understand the hydromechanical and physiological adaptation of this lake population. zse.pensoft.net 8 Acknowledgements We would like to thank Hayrullah Soylemez and Ali Haydar Kapkac, the members of the Underwater Team, and the members of the Public Security Boat Team of the Van Gendarmerie Command, for their help in diving for fish sampling. Many thanks to Saygun Dura, who photo- graphed the fish in its habitat and allowed us to use these photographs. We would like to thank Jorg Freyhof (Ber- lin) and Matthias Geiger (Bonn) for sharing the mtDNA COI sequences of O. ercisianus from the lake habitat with us. We would like to thank Harun Aydin (Ankara) for his help in understanding the microbialite hydrology. References Brinsmead J, Fox MG (2002) Morphological variation between lake- and stream-dwelling rock bass and pumpkinseed popu- lations. Journal of Fish Biology 61(6): 1619-1638. https://doi. org/10.1111/j.1095-8649.2002 .tb02502.x Costa-Pereira R, Araujo MS, Paiva F, Tavares LER (2016) Functional morphology of the tetra fish Astyanax lacustris differs between di- vergent habitats in the Pantanal wetlands. Journal of Fish Biology 89(2): 1450-1458. https://doi.org/10.1111/jfb.13026 Degens ET, Wong HT, Kempe S, Kurtman FA (1984) Geological Study of Lake Van, Eastern Turkey. Geol Rundsch 73(2): 701—734. https:// doi.org/10.1007/BF01824978 Dunn NR, O’Brien LK, Burridge CP, Closs GP (2020) Morphological convergence and divergence in Galaxias fishes in Lentic and Lotic habitats. Diversity 12(183): 1-25. https://doi.org/10.3390/d12050183 Edgar RC (2004) MUSCLE: Multiple sequence alignment with high ac- curacy and high throughput. Nucleic acids research 32: 1792-1797. https://doi.org/10.1093/nar/gkh340 Elp M, Atici AA, Sen F, Duyar HA (2016) Distribution of fish species in the Van Lake basin. Yuztincii Yil Universitesi Tarim Bilimleri Der- gisi 26(4): 563-568. https://doi.org/10.29133/yyutbd.282808 Franssen NR, Stewart LK, Schaefer JF (2013) Morphological diver- gence and flow-induced phenotypic plasticity in a native fish from anthropogenically altered stream habitats. Ecology and Evolution 3(14): 4648-4657. https://doi.org/10.1002/ece3.842 Freyhof J, Geiger MF, Golzarianpour K, Patimar R (2016) Sasanidus, a new generic name for Noemacheilus kermanshahensis Banares- cu & Nalbant, with discussion of /lamnemacheilus and Schistura (Teleostei; Nemacheilidae). Zootaxa 4107(1): 65-80. https://doi. org/10.11646/zootaxa.4107.1.3 Freyhof J, Bay¢gelebi E, Geiger MF (2018) Review of the genus Cobitis in the Middle East, with the description of eight new species (Tele- ostei: Cobitidae). Zootaxa: 4535(1): 1-75. https://doi.org/10.11646/ zootaxa.4535.1.1 Geiger MF, Herder F, Monaghan MT, Almada V, Barbieri R, Bariche M, Berrebi P, Bohlen J, Casal-Lopez M, Delmastro GB, Denys GPJ, Dettai A, Doadrio I, Kalogianni E, Karst H, Kottelat M, Kovaci¢ M, Laporte M, Lorenzoni M, Marti¢ Z, Ozulus M, Perdices A, Perea S, Persat H, Porcelotti S, Puzzi C, Robalo J, Sanda R, Schneider M, Slechtova V, Stoumboudi M, Walter S, Freyhof J (2014) Spa- tial heterogeneity in the Mediterranean Biodiversity Hotspot affects zse.pensoft.net Akkus, M. et al.: Microbialite-associated freshwater fish in Lake Van barcoding accuracy of its freshwater fishes. Molecular Ecology Re- sources 14: 1210-1221. https://doi.org/10.1111/1755-0998.12257 Haas TC, Blum MJ, Heins DC (2010) Morphological responses of a stream fish to water impoundment. Biology Letters 6(6): 803-806. https://doi.org/10.1098/rsbl.2010.0401 Ivanova NV, Zemlak TS, Hanner RH, Hebert PDN (2007) Universal primer cocktails for fish DNA barcoding. Molecular Ecology Notes 7: 544-548. https://doi.org/10.1111/).1471-8286.2007.01748.x Jones BE, Grant WD, Duckworth AW, Owenson GG (1998) Microbial diversity of soda lakes. Extremophiles 2(3): 191—200. https://doi. org/10.1007/s007920050060 Kaya C (2020) The first record and origin of Salmo trutta populations established in the Upper Tigris River and Lake Van Basin (Teleostei: Salmonidae). Journal of Anatolian Environmental and Animal Sci- ence 5(3): 366-372. https://doi.org/10.35229/jaes.777575 Kempe S, Kazmierczak J (2011) Soda Lakes. In: Reitner J, Thiel V (Eds) Encyclopedia of Geobiology. Encyclopedia of Earth Sciences Series. Springer, Dordrecht, 956 pp. https://doi.org/10.1007/978-1- 4020-9212-1 191 Kempe S, Kazmierczak J, Landmann G, Konuk T, Reimer A, Lipp A (1991) Largest known microbialites discovered in Lake Van, Tur- key. Nature 349(6310): 605-608. https://doi.org/10.1038/349605a0 Kottelat M, Freyhof J (2007) Handbook of European Freshwater Fishes. Kottelat, Cornol and Freyhof, Berlin, 646 pp. Kumar S, Stecher G, Li M, Knyaz C, Tamura K (2018) MEGA X: Mo- lecular Evolutionary Genetics Analysis across computing platforms. Molecular Biology and Evolution 35(6):1547—1549. https://doi. org/10.1093/molbev/msy096 Langerhans RB (2008) Predictability of phenotypic differentiation across flow regimes in fishes. Integrative and Comparative Biology 48(6): 750-768. https://do1.org/10.1093/icb/icn092 Langerhans RB, Layman CA, Langerhans AK, DeWitt TJ (2003) Hab- itat-associated morphological divergence in two Neotropical fish species. Biological Journal of the Linnean Society 80(4): 689-698. https://do1.org/10.1111/j.1095-8312.2003.00266.x McGuigan K, Franklin CE, Moritz C, Blows MW (2003) Adaptation of rainbow fish to lake and stream habitats. Evolution 57:104—-118. https://doi.org/10.1111/j.0014-3820.2003 tb00219.x McLaughlin RL, Noakes DLG (1998) Going against the flow: an ex- aminationof the propulsive movements made by young brook trout in streams. Canadian Journal of Fisheries and Aquatic Sciences 55: 853-860. https://do1.org/10.1139/f97-308 Perazzo GX, Corréa F, Salzburger W, Gava A (2019) Morphological dif- ferences between an artificial lentic and adjacent lotic environments in a characid species. Reviews in Fish Biology and Fisheries 29: 935-949. https://doi.org/10.1007/s11160-019-09582-y Pint: DL (2014) Soda Lakes. In: Gargaud M, Irvine WM, Amils R, Cleaves HJ, Pinti D, Quintanilla JC, Viso M (Eds) Encyclopedia of Astrobiology. Springer, Berlin, Heidelberg, 1853 pp. https://doi. org/10.1007/978-3-642-27833-4 1455-6 Reimer A, Landmann G, Kempe S (2009) Lake Van, Eastern Anatolia, Hydrochemistry and History. Aquatic Geochemistry 15: 195-222. https://doi.org/10.1007/s10498-008-9049-9 Sart M (2008) Threatened fishes of the world: Chalcalburnus tarichi (Pallas 1811) (Cyprinidae) living in the highly alkaline Lake Van, Turkey. Environmental Biology of Fishes 81: 21-23. https://doi. org/10.1007/s10641-006-9154-9 Zoosyst. Evol. 97 (1) 2021, 181-189 Schagerl M, Burian A (2016) The Ecology of African Soda Lakes: Driven by Variable and Extreme Conditions. In: Schagerl M (Ed.) Soda Lakes of East Africa. Springer, Cham, 295-320. https://doi. org/10.1007/978-3-319-28622-8 12 Swain DP, Holtby LB (1989) Differences in morphology and be- haviour between juvenile coho salmon (Oncorhynchus kisutch) rearing in a lake and in its tributary stream. Canadian Journal of Fisheries and Aquatic Sciences 46: 1406-1414. https://doi. org/10.1139/f89-180 Turan D, Kaya C, Kalayci G, Baycelebi E, Aksu I (2019) Oxynoemachei- lus cemali, a new species of stone loach (Teleostei: Nemacheilidae) from the Coruh River drainage, Turkey. Journal of Fish Biology 94(3): 458-468. https://doi.org/10.1111/jfb.13909 Webb PW (1984) Body form, locomotion, and foraging in aquatic ver- tebrates. American Zoologist 24: 107-120. https://doi.org/10.1093/ icb/24.1.107 Yosgurtguoglu B, Kaya C, Geiger MF, Freyhof J (2020) Revision of the genus Seminemacheilus, with the description of three new species (Teleostei: Nemacheilidae). Zootaxa 4802(3): 477-501. https://doi. org/10.11646/zootaxa.4802.3.5 Zar JH (2010) ‘Biostatistical Analysis’. Prentice Hall International: Up- per Saddle River, NJ, USA. zse.pensoft.net