Verhandlungen der Gesellschaft für Ichthyologie, Band 5, 71-77 Reduction of scales and head canals in Pomatoschistus canestrinii (Ninni, 1883) (Teleostei, Gobiidae) Reduktion von Schuppen und Kopfkanälen bei Pomatoschistus canestrinii (Ninni, 1883) (Teleostei, Gobiidae) Björn Stelbrink and Jörg Freyhof Leibniz-Institute of Freshwater Ecology and Inland Fisheries, Department of Biology and Ecology of Fishes, Müggelseedamm 310, D-12587 Berlin, Germany; GreatWhiteSharks@t-online.de Summary: Pomatoschistus canestrinii from freshwater Lake Skadar (Montenegro) have reduced squamation and head canals, while the brackish water population from the River Krka (Croatia) and two additional freshwater populations from the River Ricica (Croatia) and the Bacinska Lake system (Croatia) show a complete or almost complete squamation and head canal system. In regard to the distribution patterns of the head canals as well as head pores, partial reductions could be pointed out for Bacinska as well as Ricica individuals, while the largest reductive range of variation up to the complete loss of the canal system could be noticed within the Lake Skadar population. The ecological factors, which lead to these reductions, remain unresolved. Key words: Adriatic Sea, scale reduction, head canal reduction, Gobiidae, Pomatoschistus Zusammenfassung: Pomatoschistus canestrinii aus dem Skutarisee (Montenegro) haben eine reduzierte Beschuppung sowie reduzierte Kopfkanäle, während die Brackwasser-Population aus dem Fluss Krka (Kroatien) sowie zwei zusätzliche Süßwasser-Populationen aus dem Fluss Ricica (Kroatien) und dem Bacinska Seensystem (Kroatien) eine komplette Beschuppung bzw. ein komplettes oder nahezu komplettes Kopfkanalsystem aufweisen. In Bezug auf das Verteilungsmuster der Kopfkanäle und -poren können teilweise Reduktionen bei den Bacinska- sowie Ricica-Individuen aufgezeigt werden, während die größte Bandbreite an Reduktionen, bis hin zum vollständigen Verlust des Kanalsystems, innerhalb der Skutariesee-Population beobachtet werden kann. Die ökologischen Faktoren, die zu diesen Reduktionen führen, bleiben allerdings ungeklärt. Schlüsselwörter: Adria, Schuppenreduktion, Kopfkanalreduktion, Gobiidae, Pomatoschistus 1. Introduction European freshwater sand-gobies are traditionally subdivided into the genera Economidichthys, Knipowitschia and Pomatoschistus (Economidis and Miller 1990). The phylogeny of sand-gobies is considered to be poorly understood and traditional genera might form paraphyletic assemblages, originating from multiple independent origins by marine ancestors (Penzo et al. 1998). While Economidichthys is distinguished from the other two genera by an unique perianal organ and transverse supraorbital papillae rows (Miller 2004), Knipowitschia and Pomatoschistus are delimited from each other by the length of the anterior oculoscapular canal, which is extending to the snout, ending at pore σ in Pomatoschistus (vs. ending before pore σ in Knipowitschia) (fig. 1). The posterior oculoscapular canal is present in Pomatoschistus, whereas in other sand gobies Verh. Ges. Ichthyol. Bd. 5, 2006 71
Figs. 1a, b: Head lateral-line canal pores (Greek letters) and sensory papillae in dorsal view (from Miller 2004) of a Pomatoschistus canestrinii and b Knipowitschia caucasica. AN, PN, anterior and posterior nostrils. Abb. 1a, b: Kopfkanalporen (griechische Buchstaben) und sensorische Papillen in Dorsalansicht (aus Miller 2004) von a Pomatoschistus canestrinii und b Knipowitschia caucasica. AN, PN, vordere und hintere Nasenöffnungen. it is usually absent (Miller 2004). However, it is well known that sand gobies of the genus Knipowitschia might reduce their head canals what makes their generic identification impossible by the characters listed above. Because no alternative generic concept is available, freshwater species with reduced head canals are generally placed in the genus Knipowitschia (Economidis and Miller 1990, Ahnelt 1991). Fourteen species of estuarine and freshwater gobies from the Caspian, Black and Mediterranean Sea basins are included in the genus Knipowitschia. Within Knipowitschia, there is a group of species with complete squamation along lateral series until the caudal fin base and presence of complete or almost complete cephalic lateral line canals (K. bergi, K. cameliae, K. caucasica, K. goerneri, K. iljini, K. longecaudata, and K. panizzae) (Miller 2004). All these species are euryhaline inhabitants of coastal lagoons and river estuaries. A second group of species is characterized by incomplete or absent squamation along lateral series and/or incomplete or absent cephalic lateral line canals (K. croatica, K. ephesi, K. mermere, K. milleri, K. punctatissima, K. thessala) (Miller 2004). All these species are restricted to freshwater habitats. While both species of Economidichthys exhibit a reduced squamation and no head canals, all species of Pomatoschistus have a complete squamation and complete cephalic lateral-line canals (Miller 2004). Whereas Economidichthys inhabit freshwaters only (Miller 2004), P. microps is the only amphidromous species entering freshwaters during seasonal migrations (Miller 2004, Riede 2004). All other Pomatoschistus are resident inhabitants of marine or at least brackish waters (Miller 2004). Miller (2004) already speculated that reduction of squamation and head canals could be correlated with the invasion of sand gobies into freshwater. To test this hypothesis, a sand goby should be studied, which represents a resident inhabitant of estuarine and freshwater habitats. In Europe two gobies come into question: Knipowitschia caucasica and Pomatoschistus canestrinii. Knipowitschia caucasica individuals seem to be monomorphic or at least less variable than the other species in freshand brackish waters, representing the complete pattern of squamation and head canals (Miller 2004). Pomatoschistus canestrinii is a euryhaline goby, well known from estuaries and lagoons in 72
the northern Adriatic as well as along the Croatian coast (Miller 2004). While only brackish and marine populations were known, an introduced freshwater population was first recorded by Freyhof (1998) from Italian Lake Trasimeno, demonstrating that this species is able to live in pure freshwater. Well known, but misidentified (Ivanovic 1973), freshwater gobies from Lake Skadar basin were recently identified as freshwater population of P. canestrinii by Bohlen et al. (2003). During an expedition to Croatia and Montenegro this species could be collected in one estuarine and three freshwater habitats. Brackish and freshwater populations of the goby P. canestrinii were examined to verify possible reductions in this species and to evaluate the potential correlation between this process of reduction and the invasion into freshwater habitats. 2. Material and methods All fish were caught using a fine mesh beach seine or a DEKA 3000 portable electroshocker and preserved directly in 5 % formaldehyde, and later transferred to 70 % ethanol for storage. Abbreviation FSJF used denotes the fish collection of Jörg Freyhof (Fischsammlung Jörg Freyhof). River Krka individuals (FSJF 1090) represent the brackish water reference population for the estuarine type; the other locations represent freshwater populations. Material examined: P. canestrinii. FSJF 1090, n = 20; 10 males, 10 females, 27.39-34.55 mm SL; Croatia: River Krka at uppermost part of Prokljansko Jezero at Skradin; Nina Bogutskaya & Jörg Freyhof. FSJF 1074, n = 13; 2 males, 11 females, 20.05-30.08 mm SL; Croatia: reservoir at River Ricica at village Stikada, 44 19.228 N 15 46.940 E; Nina Bogutskaya & Jörg Freyhof. FSJF 1088, 1 male, 34.41 mm SL; Croatia: Lake Bacinska at village Bacina; Nina Bogutskaya & Jörg Freyhof. FSJF Verh. Ges. Ichthyol. Bd. 5, 2006 1048, n = 5; 1 male, 4 females, 26.10-30.18 mm SL; Montenegro: Lake Malo Blato at Begova Glavica; Primoz Zupancic & Jörg Freyhof. FSJF 1096, n = 40; 20 males, 20 females, 23.44-31.95 mm SL; Montenegro: backwater of River Moraèa, west of bridge between Golubovci and Vukovici, right river side; Primoz Zupancic & Jörg Freyhof. Pomatoschistus canestrinii (fig. 2) is distinguished from other Adriatic goby species in morphological and coloration characters, using the presence or reduction and loss, respectively, of head canals and squamation as well as the body coloration, e.g. numbers of vertical bars. Males and females on their part are distinguished in body coloration. Whereas males show a prominent black head with yellowwhitish neck blotches as well as a dark grey pelvic disc, 5-6 dark vertical bars and spots, the females are characterized by the absence of bars, a whitish head and pelvic disc as well as fewer spots (Miller 2004). The body squamation as well as the head canal and pore systems were compared with the species habitus shown by Miller (2004) and cephalic lateral-line canal description and canal and pore definition, both using a binocular microscope. Within the head canal system three major structures are crucial for the taxonomic discrimination of sand-gobies: the anterior oculoscapular canal, including pores σ, λ, κ, α, ω and ρ, furthermore the posterior oculoscapular canal, including the two pores ρ 1 and ρ 2 and finally the preopercular canal with pores γ, δ and ε (for better understanding see fig. 4). 3. Results The population from brackish waters of lower River Krka displays full body squamation and a lack of scales on the predorsal body (fig. 3). The same squamation type could be noticed within the River Ricica 73
Fig. 2: Pomatoschistus canestrinii, male, Montenegro, River Moraca. Abb. 2: Pomatoschistus canestrinii, Männchen, Montenegro, Fluss Moraca. Fig. 3: Populations of P. canestrinii examined. Red dots denote study sites. Red marks on fish drawings show dominant squamation type. Circular charts show the head canal type distribution pattern within study sites. Outline map shows distribution of P. canestrinii, following Miller (2004). Abb. 3: Untersuchte Populationen von P. canestrinii. Rote Kreise zeigen Lage der Fundorte. Die rote Markierung auf den Fischzeichnungen zeigt das dominante Ausmaß der Beschuppung der Populationen. Die Kreisdiagramme zeigen die in der jeweiligen Population dominante Ausprägung der Kopfkanäle. Die Übersichtskarte zeigt die Verbreitung von P. canestrinii nach Miller (2004). 74
and Lake Bacinska freshwater individuals. Lake Skadar population is distinguished from the estuarine type by a reduction of the scales up to a small region underneath the pectoral fins (fig. 3). Within the P. canestrinii populations 36 head canal types could be observed. Similar forms were summarized into five arbitrary types (fig. 4) to get better information about the distribution pattern. Type I represents the estuarine type with a complete or almost complete anterior oculoscapular canal, types II-V represents the arbitrarily defined gradual reduction of the anterior oculoscapular canal and preopercular canals and pores. The posterior oculoscapular canal is absent in all specimens examined. Type I can be noticed within the estuarine population and the single Lake Bacinska individual. The Ricica individuals are almost homogenous and possess the type I (n = 4; 4 females) as well as type II (n = 9; 2 males, 7 females). The Lake Skadar population is represented by type II (n = 6; 5 males, 1 female), type III (n = 15; 9 males, 6 females), type IV (n = 7; 3 males, 4 females) and type V (n = 17; 4 males, 13 females). Thus, the largest variation can be observed within the Lake Skadar population, while the other populations more or less belong to one head canal system type (fig. 4). Fig. 4: Head canal types. Greek letterings denote pores, dashed lines denote canals, solid lines denote open canals (furrows). Denotation of pores and canals after Miller (2004). AN, PN, anterior and posterior nostrils. Undefined pores could not be classified into the intraspecific distribution pattern. Abb. 4: Kopfkanaltypen. Griechische Buchstaben bezeichnen Poren, gestrichelte Linien den Kanalverlauf, durchgezogene Linien offene Kanäle (Furchen). Bezeichnung der Poren und Kanäle nach Miller (2004). AN, PN, vordere und hintere Nasenöffnungen. Unbeschriftete Poren konnten nicht in das intraspezifische Verteilungsmuster eingeordnet werden. Verh. Ges. Ichthyol. Bd. 5, 2006 75
4. Discussion The scale and head canal reduction in P. canestrinii from Lake Skadar is remarkable, not only being the single examined population with this trait, but also being a homogenous, monomorphic population, regarding this character. First, one can ask why the other freshwater populations from the northern regions do not display these scale reductions, but look like the brackish water populations. And secondly, is this reduction of scales due to the Lake Skadar abiotic and biotic factors? If the reduction of squamation and head canals is related to the invasion of freshwater habitiats, the time passed since the colonization of the freshwater habitats might be crucial for the understanding of this process, which might be speculated to be an adaptive one. In this case, it could be suspected that Lake Skadar P. canestrinii are remarkable longer isolated from the coast than the other examined populations. The situation in P. canestrinii is somehow similar to the reduction of lateral plates in sticklebacks, Gasterosteus aculeatus. There is a wide range of studies, dealing with plate reduction in G. aculeatus with different points of view, whose results and consequential discussions could be compared with the results from the present study. For sticklebacks it is supposed that lateral plates mainly acting as armors against toothed fishes and increasing the predator s handling time. Vice versa, the reduction of plates is therefore due to a lower predatory pressure (e.g. Fernandez et al. 2000, Reimchen 2000). The same might be also suspected for P. canestrinii. Further limiting factors, which are mentioned by these and other authors (salinity, temperature and calcium concentration), may also be important for the magnitude of plates in sticklebacks and scales in P. canestrinii, respectively. The function of the cephalic lateral-line system can be devided into prey detection [ ], rheotaxis [ ], schooling [ ], intraspecific communication [ ] and identification and localization of wave sources [ ] (reviewed by Engelmann et al. 2002). Ahnelt (2001) presumes that the varied course of the head canals in Gobius ater and G. paganellus may due to the different habitats they are living in. Therefore, the reductions in the lateral head canal system in the Lake Skadar population could display also a habitat-specific adaptation possibly supported by the recently described speciesspecific distribution pattern of P. canestrinii in the Venice Lagoon (Malavasi et al. 2005), conversely, a complete lateral head canal system may be more advantageous in brackish or marine ecosystems for P. canestrinii. It could be supposed that a lower predation pressure might also force the reduction of head canals as gobies may detect a predator also by wave sources identification and localization. Kovacic (2005) also reveals reduction patterns of the head canal system as well as the squamation among P. canestrinii populations from several places along the Croatian coast. His findings in reductions of the cephalic lateral-line system are quite similar to those of our study, even though the specimens examined by Kovacic (2005) show a lower variability as well as a lower degree of reduction. Also, the pore ω, which was not found by Kovacic (2005), could be recorded in specimens from Lake Skadar and in some Ricica individuals. Hence, by the hypothesized transformation series in fig. 4 we assume that a canal reduction begins with a shortening/contraction of the mentioned canal and thus with an adjacent displacement as well as a new development of pores (e.g. pore ω in this case), finishing with the loss of the last parts of the canals and the pores themselves, respectively. By this, a doubled interorbital canal could be noticed in some specimens. 76
Another morphological trait, considered by Kovacic (2005), is the development of the squamation. Among the study sites two character states were noticed by Kovacic (2005), a continuous squamation along the lateral midline and a squamation reduced to axillar and caudal areas. Whereas the complete squamation type is also observed within the present study, the reduced squamation type with two scaled areas was not observed. In Lake Skadar, an axillar scale patch occurred, but scales were always absent from the caudal peduncle. Future investigations are necessary to clarify the biogeography of P. canestrinii and the timing of the invasion of different freshwater bodies. The questions, which ecological factors lead to the observed reductions, remain unresolved. Laboratory experiments with naked and scaled P. canestrinii could be performed to test if scaled P. canestrinii with a full set of head canals are less vulnerable to predator pressure than naked P. canestrinii or if predators need a longer handling time preying on scaled P. canestrinii. Acknowledgements We are thankful to Primoz Zupancic ` ` (Ljubljana) for his help during field work. Literature Ahnelt, H. 1991. A new species of Knipowitschia (Teleostei: Gobiidae) from Corfu, western Greece. Ichthyol. Explor. Freshwaters 2, 265-272. Ahnelt, H. 2001. Two Mediterranean gobiid fishes with an unusual cephalic lateral line system. Cybium 25, 261-267. Bohlen, J., Šlechtová, V., Šanda, R., Kalous, L., Freyhof, J. Vukic, J., Mrdak, D. 2003. Cobitis ohridana and Barbatula zetensis in the River Moraca basin, Montenegro: Distribution, habitat, population structure and conservation needs. Folia Biol. 51, Suppl., 147-153. Economidis, P.S., Miller, P.J. 1990. Systematics of freshwater gobies from Greece (Teleostei: Gobiidae). J. Zool., London 221, 125-170. Engelmann, J., Hanke, W., Bleckmann, H. 2002. Lateral line reception in still and running water. J. Comp. Physiol. A 188, 513-526. Fernandez, C., Hermida, M., Amaro, R., San Miguel, E. 2000. Lateral plate variation in Galician stickleback populations in the rivers Miño and Limia, NW Spain. Behaviour 137, 96-979. Freyhof, J. 1998. First record of Pomatoschistus canestrinii (Ninni, 1883) in Lake Trasimeno. Riv. Idrobiol. 37, 107-108. Ivanovic, B.M. 1973. Ichthyofauna of Skadar Lake. Institution for Biological and Medical Research in Montenegro, Biological Station, Titograd. Kovacic, M. 2005. Morphological variability of Pomatoschistus canestrinii (Gobiidae), with the reduction of squamation and head canals. Cybium 29, 373-379. Malavasi, S, Franco, A., Fiorin, R., Franzoi, P., Torricelli, P, Mainardi, D. 2005. The shallow water gobiid assemblage of the Venice Lagoon: abundance, seasonal variation and habitat partitioning. J. Fish Biol. 67, 146-165. Miller, P.J. 2004 (ed.). The Freshwater Fishes of Europe. Vol. 8/II, Gobiidae 2. Aula-Verlag, Wiebelsheim. Penzo, E., Gandolfi, G., Bargelloni, L., Colombo, L., Patarnello, T. 1998. Messinian salinity crisis and the origin of freshwater lifestyle in western Mediterranean gobies. Mol. Biol. Evol. 15, 1472-1480. Reimchen, T.E. 2000. Predator handling failures of lateral plate morphs in Gasterosteus aculeatus: Functional implications for the ancestral plate condition. Behaviour, 137, 1081-1096. Riede, K. 2004. Global register of migratory species from global to regional scales. Final Report of the R&D-Projekt 808 05 081. Federal Agency for Nature Conservation, Bonn, Germany. Verh. Ges. Ichthyol. Bd. 5, 2006 77
78