Verhandlungen der Gesellschaft für Ichthyologie Band 5, 2006, 139-163

Preliminary checklist of sailfin silversides (Teleostei: Telmatherinidae) in the Malili Lakes of (), with a synopsis of systematics and threats

Vorläufige Checkliste der Sonnenstrahlfische (Teleostei: Telmatherinidae) der Maliliseen Sulawesis (Indonesien), mit einer Übersicht zu Systematik und Bedrohung

Fabian Herder1, Julia Schwarzer1, Jobst Pfaender1, Renny K. Hadiaty2, and Ulrich K. Schliewen3. 1 Sektion Ichthyologie, Zoologisches Forschungsmuseum Alexander Koenig, Adenauer- allee 160, D-53113 Bonn, Germany; [email protected] (corresponding author) 2 Ichthyology Laboratory, Division of Zoology, Research Center for Biology, Indonesian Institute of Sciences (LIPI), Jl. Raya Bogor Km 46, Cibinong 16911, Indonesia 3 Department of Ichthyology, Bavarian State Collection of Zoology (ZSM), Münchhausenstr. 21, D-81247 München, Germany

Summary: The ancient Malili Lakes in Sulawesi harbor an adaptive radiation of sailfin silversides (Teleostei: : Telmatherinidae). This radiation is characterized by local endemism either to basins of Lakes Towuti, Mahalona and Lontoa (Wawontoa) or to that of Lake Matano, and by spectacular male color polymorphisms. Based on recent surveys, we report a previously unde- tected diversity of steam-dwelling Telmatherinidae, inhabiting a wide range of the lake system’s rivers and streams and additional adjacent catchments. We provide a preliminary checklist of Telm- atherinidae in the Malili Lakes system, including surrounding catchments, and add records of three yet undescribed lacustrine from Lake Matano. Distribution patterns of Lake Matano’s Telmatherina are provided in detail, based on eight sampling locations around the lake’s shoreline. Moreover, we record an increased number of color morphs in five known Telmatherina and summarize color polymorphisms of the species-flock. As this radiation is used as model system to address questions concerning speciation and maintenance of color polymorphisms, we provide a summary of taxonomic, systematic, ethological and evolutionary biological knowledge regarding telmatheri- nids, give a brief overview about the environmental settings, and discuss the serious threats endan- gering this endemic ichthyofauna.

Key words: Telmatherinidae, adaptive radiation, Malili Lakes system, Sulawesi, diversity, threats

Zusammenfassung: Die Maliliseen Sulawesis beherbergen eine adaptive Radiation von Sonnen- strahlfischen (Teleostei: Atheriniformes: Telmatherinidae). Diese Radiation zeichnet sich durch lokalendemische Verbreitungsmuster einerseits in den Seen Towuti, Mahalona und Lontoa (Wa- wontoa) oder andererseits im Matanosee, sowie durch spektakuläre Farbpolymorphismen aus. Ba- sierend auf kürzlich erfolgten Felderfassungen stellen wir eine bislang unbekannte Diversität bach- bewohnender Telmatheriniden vor, die in weiten Bereichen der Flüssen und Bächen des Seensys- tems sowie angrenzenden Einzugsgebieten verbreitet ist. Wir stellen eine vorläufige Checkliste der Telmatheriniden des Maliliseen-Systems einschließlich umgebender Einzugsgebiete auf und fügen Nachweise von drei bislang unbeschriebenen lakustrinen Telmatherina aus dem Matanosee hinzu. Verbreitungsmuster der Manatosee-Telmatherina werden detailliert aufgezeigt, basierend auf acht Sammelstellen rings um die Uferzone des Sees. Außerdem weisen wir eine höhere Anzahl Farb-

Verh. Ges. Ichthyol. Bd. 5, 2006 139 morphen für fünf bekannte Telmatherina nach und fassen die Farbpolymorphismen des Artenschwar- mes zusammen. Da diese Radiation als Modellsystem dient um Fragen zu Artbildung und Aufrech- terhaltung von Farbpolymorphismen nachzugehen, geben wir eine Zusammenfassung über den taxonomischen, systematischen, verhaltensbiologischen und evolutionsbiologischen Forschungs- stand über Telmatheriniden sowie einen kurzen Überblick über die naturräumlichen Bedingungen und diskutieren die ernsthaften Bedrohungen, die diese endemische Fischfauna gefährden.

Schlüsselwörter: Telmatherinidae, Adaptive Radiation, Maliliseen, Sulawesi, Diversität, Bedro- hung

1. Introduction examinations and a sophisticated concept of recognizing “species” within this ob- Over the last 109 years a diversity of small, viously still radiating flock are required. atheriniform freshwater has been described from the island of Sulawesi (In- 1.1. The Malili Lakes system donesia), accounted to the family of sailfin silversides (Atheriniformes: Telmatherin- The Malili Lakes system is located in Cen- idae). Only one out of the island’s 16 spe- tral Sulawesi (the former Celebes) (fig. 1), cies is found in streams of its south-west- consisting of five interconnected lakes. The ern branch, the remaining 15 are endemic three major lakes Matano, Mahalona and to the Malili Lakes system in Central Sula- Towuti are large, deep ancient lakes of tec- wesi (Kottelat 1990a, 1991). The chance to tonic origin (Abendanon 1915a, b, Brooks study evolution of organismic diversity 1950, Haffner et al. 2001). Lake Matano is within an outstanding, compact but diverse with 590 m the deepest lake in Southeast model system characterized by different Asia and the eighth deepest lake in the world degrees of isolation, has attracted recently (Hutchinson 1957). Geologically, it is a tec- several research groups to this system. Apart tonic graben lake of the still active Matano from sailfin silversides, the endemic inver- fault zone (Ahmad 1977). tebrate radiations of crustaceans (atyid Lakes Matano, Mahalona and Towuti are shrimps and parathelphusid crabs) and pa- arranged along a cascading chain with con- chychilid snails (von Rintelen and Glau- necting rivers. Lake Matano is a steep-si- brecht 2003; von Rintelen and Glaubrecht ded graben lake (Ahmad 1977), located 396 2005; von Rintelen et al. 2004) as well as m above sea level (fig. 1). It is drained at its the predominantly endemic diversity of dia- eastern corner by the extremely steep toms (Bramburger et al. 2004, 2006) are River Petea to L. Mahalona, descending 72 m under investigation. of altitude along the river’s total length of Aim of the present paper is to (1) summar- about 9.5 km (Abendanon 1915a, Kottelat ize present knowledge about telmatherinid 1991, own observations). According to our , systematics and the Malili Lakes own ground checks and all accessible topo- system and (2) to provide an overview about graphic maps (1: 50.000, Lembar 2213-13, major patterns of diversity in the Malili 34) there is no major waterfall in River Lakes sailfin-silversides radiation. It is in- Petea. The river is located within a very tended as a guide to upcoming work dea- steep valley and descends along a sequence ling with evolutionary processes of this of rapids, riffles, small falls, pools and back- adaptive radiation, and to drawing attention waters until about 1 km ahead of L. Maha- to serious threats endangering this unique lona, where it becomes comparatively deep diversity. It is not aimed to serve taxono- and calm. After Lake Mahalona, the flow mic purposes, for which more detailed continues as the overall wide and slow-flow-

140 ing River Tominanga which has only one flow, Abendanon (1915b) indicated a water- series of rapids and covers a 17 m drop in fall. However, there is no indication in the elevation before reaching L. Towuti after map pointing to anything like a major water- about 8.1 km. With approximately 560 km2 fall with a drop of dozens of meters as re- of surface area, L. Towuti is the largest lake cently claimed (Roy et al. 2004). in Sulawesi (Haffner et al. 2001), but is with Brooks (1950) reviewed knowledge about 203 m less deep than L. Matano. It is drained the Malili Lakes and their species flocks and at its south-western corner to the Gulf of speculated about reasons for the patterns Bone by River Larona (= R. Malili). of their organism’s distribution. Based on The two small satellite lakes Lontoa (or Abendanon’s map, he focussed on the sep- Wawontoa; local spelling Lantoa) and Masa- arating effect of waterfalls along River pi are probably not of tectonic origin Petea, stating “…the relatively short Patea (Abendanon 1915a). According to Abenda- River which drains into Lake Mahalona de- non (1915b), the outflow of L. Lontoa is scends 72 m., part of this drop occurring located at its northern corner, draining the as waterfall.” (Brooks 1950, p. 167). lake to River Tominanga. In contrast, mod- Recent investigations of the lake’s water ern topographic maps based on aerial photo- chemistry (Giesen et al. 1991) and L. graphs show a direct connection from its Matano’s limnology (Haffner et al. 2001) southern end to L. Towuti by a small stream underpinned that all five lakes are ultraoli- (1: 50.000, Lembar 2213-13, 34). Lake Masa- gotrophic, with different levels of deficiency pi has a crater lake appearance, although in various trace elements. there is no evidence for volcanism in the area (Giesen et al. 1991). It drains to River 1.3. Telmatherinid discovery, syste- Larona, without a direct connection to the matics and evolution large lakes. Boulenger (1897) described the first sailfin 1.2. Discovery of the Lakes silverside as , and placed it in Atherinidae. Both syntypes originate Fritz and Paul Sarasin, two Swiss geologists from the collection by P. and F. Sarasin in and naturalists, were the first “western” visi- 1896 at (see Kottelat and Sutt- tors who explored the greater area of Lake ner 1988). Based on material gathered in Matano (as “Matana”) and Towuti in early 1909 by Abendanon at Soroako, Lake Mata- 1896 (Sarasin 1886). They provided a pre- no, Weber (1913) described Telmatherina liminary map and were the first to collect abendanoni and mentioned two additional telmatherinids, which later were reported specimens of Telmatherina from Lake To- by Boulenger (1897). The Dutch geographer wuti that differed from T. celebensis in their and geologist E. C. Abendanon explored number of anal fin rays; these two speci- these two lakes in 1909 and discovered Lakes mens were referred to as “Bonti” by local Mahalona, Lontoa and Masapi (Abendanon inhabitants. In 1922, Weber and de Beau- 1915a, b). In addition, he provided the first fort described Telmatherina bonti based on hydrological data of the lakes as well as these two specimens. detailed maps of the complete lake system. Aurich (1935) revised telmatherinids based Noteworthy, Abendanon precisely draw the on material collected by Woltereck’s course of River Petea (as “Patea”), which “Wallacea Expedition”, redescribed Telm- delimits for most of its stretch the north- atherina bonti, T. celebensis and T. abendanoni western extreme of the area mapped in based on detailed morphometric and meri- Abendanon (1915b). In the middle of its stic analyses, and described four new spe-

Verh. Ges. Ichthyol. Bd. 5, 2006 141 cies which he placed in the new Par- sal and anal fin” to distinguish Telmatherin- atherina (P. wolterecki, P. labiosa, P. striata, P. idae from Pseudomugilidae and Melano- cyanea). He also provided a key to all telm- taeniidae (Munro 1967). Kottelat (1990a, atherinid species known at that time, but did 1991) recognized and revised the family not designate a type species for . based on new material, and provided a “ten- Aurich (1935) noticed that fin coloration is tative diagnosis” (1990a, p. 228-229) of the variable in Telmatherina from Lake Towuti family, distinguishing it from Melanotaeni- (not separated to T. bonti / T. celebensis), with idae, Atherinidae, Dentatherinidae, Pseudo- yellow, bluish or red margins of first and mugilidae, Bedotiidae, Isonidae and Phal- second dorsal as well as anal fins. In addi- lostethidae, which is partly in discordance tion, he provided a drawing of a single male with Munro’s diagnosis and refers substan- specimen of a yet undescribed Telmatherina tially to osteological characters provided by from L. Matano, characterized by fleshy lips personal communication from W. Ivantsoff. (Telmatherina sp. “thicklip”; see below). Un- According to external morphology, Kottelat fortunately, most of Aurich’s material, in- (1990a) diagnosed Telmatherinidae as me- cluding all Paratherina type material, was dium to large atherinoids (=suborder Athe- destroyed in the Bavarian State Collection rinoidei within Atheriniformes) exhibiting of Zoology, Munich, Germany (ZSM) dur- strong sexual dimorphism with dull brown, ing Second World War (Kottelat 1990a, b). short-finned females and yellow, blue or red Ahl (1936) described with Telmatherina ladi- male body and fin coloration and very long gesi the first telmatherinid from outside the first and second dorsal and anal fins of Malili Lakes system. The three specimens males. available to him were collected by Ladiges Kottelat (1990a, 1991) described eight new and Arnold (without collection date) from telmatherinid species, including two of the a rivulet of Makassar’s hinterland. Ahl pro- newly erected genus Tominanga (T. aurea, T. vided a short key for other Telmatherina spe- sanguicauda) and six species of Telmatherina cies, but did not discuss to family assign- endemic to Lake Matano (T. antoniae, T. ob- ment of the genus. scura, T. opudi, T. prognatha, T. sarasinorum, T. Brooks (1950) included the Malili Lakes into wahjui). Tominanga is defined by its slender his review of “Speciation in Ancient body and shape of pelvic fins which form Lakes”, referred to the lacustrine Telm- a funnel-shaped ovipositor in females. He atherinidae as a “species flock” and dis- (Kottelat 1990a) validated the genus Par- cussed their potential origin and history as atherina from Lakes Towuti, Mahalona and a consequence of isolation by tectonic pro- Lontoa (“Wawontoa”) by designation of a cesses and limited dispersal capabilities due neotype for P. wolterecki and designation of to strongly isolating watersheds and water- P. wolterecki as type species. The genus is falls. He also discussed to possibility of in- defined by scale counts and in having two tralacustrine speciation within Lakes Towuti rows of scales below eye (vs. one in Telm- and Matano. atherina and Tominanga). However, Kottelat’s In 1958, Munro established the family Telm- redescription of Paratherina species was re- atherinidae without diagnosis (Munro 1958), stricted severely by limited material, i. e. a including the Australian Quirichthys strami- total of five juvenile specimens (2 speci- neus Whitley (currently placed in Crateroce- mens of P. cyanea, 2 of P. labiosa, and the phalus; Crowley and Ivantsoff 1992) and neotype of P. wolterecki; sizes of only 9.6- diagnosed it later using as characters “posi- 43.2 mm SL). tion of first dorsal fin relative to ventral- Kottelat’s descriptions (1990a, 1991) include fin” and “elongated fin rays in second dor- male color polymorphisms in several Telm-

142 atherinids. He reported blue/yellow male Dyer and Chernoff (1996) was restricted morphs in Telmatherina antoniae, T. celebensis, to T. ladigesi, diminishing their argumenta- T. sarasinorum and T. wahjui, and red/yellow tion to place Telmatherinini + Pseudomu- male morphs in Tominanga aurea. However, gilini together with Melanotaeniinae and Be- he pointed out (Kottelat 1990a, 1991) that dotiinae into one single family Melanotae- his review does not pretend to be complete niidae (for which, following Aarn et al. 1998, with respect to total coverage of Telmatherin- Pseudomugilidae would be the correct idae. His purpose was to make names avail- name). able (Kottelat 1991) based on the material Aarn and Ivantsoff (1997) presented a phy- available to him. Consequently, he fore- logenetic analysis of Melanotaeniidae based casted that parts of his work would have to on 46 morphological characters, including be revised in the light of upcoming further the telmatherinids Kalyptatherina helodes and work, especially with respect to L. Matano’s Telmatherina ladigesi as well as Paratherina wol- telmatherinids (Kottelat 1991) and Parathe- terecki, Telmatherina abendanoni, T. antoniae, T. rina (Kottelat 1990a). bonti, T. celebensis, T. opudi, T. sarasinorum, T. In a series of popular articles, Kottelat wahjui, Tominanga aurea and T. sanguicauda. (1989a, b, c, 1990c, 1992) provided addi- They presumed monophyletic Telmatherin- tional information about the Lake’s geo- idae and found the family together in a po- graphic settings and their freshwater fauna. lytomy with Atherinoidae, Phallostethidae, Kottelat (1989b) illustrated yellow and blue Atherinidae, Pseudomugilidae and Melano- morphs of T. celebensis. Later he (Kottelat et taeniidae to be the sistergroup of Atherin- al. 1993) included all known telmatherinids opsidae and Notocheiridae. in a guidebook, provided drawings and illus- Aarn et al. (1998) analyzed intra-telmatheri- trated the most important characters. There, nid relationships on generic level compared the first color pictures of T. obscura were to Atherinidae, Melanotaeniidae and Pseu- made available and color pictures of some domugilidae. Based on 26 morphological preserved material displayed only in black characters and essentially the same telm- and white before (Kottelat 1990a, 1991) atherinid material as Aarn and Ivantsoff were included. (1997), they found support for monophy- Saeed and Ivantsoff (1991) investigated letic Telmatherinidae with sister group re- osteological characters of Pseudomugil helo- lationship between Kalyptatherina + Telm- des with respect to Pseudomugilidae, Phal- atherina ladigesi and all Malili Lakes telm- lostethidae, Atherinidae, Notocheiridae, atherinids; the clade Telmatherina + Tominanga Isonidae, Dentatherinidae, Melanotaeniidae, + Paratherina was not resolved. Subsequently, Bedotiidae and Telmatherina celebensis, T. bon- they rediagnosed Telmatherinidae and ti and T. ladigesi as representatives of Telm- erected the genus Marosatherina for T. ladi- atherinidae. They concluded that P. helodes gesi, which they found to differ substantial- from islands Batanta and Misool off New ly in several anatomical characters from the Guinea is most closely related to Telm- known telmatherinid genera. However, the atherinidae, but sufficiently distinct to be authors state in the discussion, that mono- placed into a new genus, Kalyptatherina. phyly of Telmatherina (also without Marosa- Saeed et al. (1994) as well as Dyer and Cher- therina) is doubtful, which might be the rea- noff (1996) investigated morphological re- son for low resolution in their generic tree. lationships among atheriniform families, In addition to their phylogenetic work, Aarn supporting either family (Saeed et al. 1994) et al (1998, p. 320) mentioned the occur- or tribe (Dyer and Chernoff 1996) rank for rence of Telmatherina bonti in “creeks of the Telmatherinidae. Telmatherinid material in Matano basin and outside of the Malili lake

Verh. Ges. Ichthyol. Bd. 5, 2006 143 basin, in Sungei Masiluk, a basin adjacent upon the assumption that L. Matano is to Lake Towuti and draining eastern to the “…separated from downstream lakes by a Banda sea”, without providing reference. 72-m waterfall…” (Roy et al. 2004, p. 2), They speculated that T. bonti could turn out which is not supported by recent evidence as sister species of all other Telmatherina or (see above). Second, they state that L. Mata- even all Malili Lakes telmatherinids. no ”...has been identified as the principal Hadiaty and Wirjoatmodjo (2002) reported seeding source of endemic species within preliminary results from their field studies the Malili Lakes watershed…”, citing at Lake Matano, listing all seven described Brooks (1950) and Haffner et al. (2001). Telmatherina of the lake as well as one un- This implicates knowledge about the flock’s described species of the genus. Moreover, evolutionary history. However, none of they reported T. bonti to occur at the mouth both references provides any data on phy- of streams entering the lake at its north- logenetic relationships. Third, Roy and co- western corner. authors claim that their “…16S rDNA se- Sparks and Smith (2004) combined in a quences demonstrate common ancestry for phylogenetic analysis of Malagasy and Aus- each of the endemic genera in Lake Mata- tralasian (Melanotaenioidei) no…” (Roy et al. 2004, p. 7). This con- DNA sequence data from nuclear and mi- clusion appears premature because they did tochondrial genes with morphological data. not include any species from outside L. In this analysis, their single telmatherinid, Matano in their analysis, although all four Marosatherina ladigesi, was nested within genera are also present outside L. Matano. Pseudomugilidae, comprising together the As a consequence, we cannot accept the sister group to Melanotaeniidae. They con- principal line of evidence presented by Roy clude that Telmatherinidae should be in- et al. (2004) although we do not disagree cluded into Pseudomugilidae. However, with their principal conclusion, i.e. that sail- missing representation of more than 94% fin silversides of L. Matano represent a case of the described telmatherinid species as of adaptive radiation. well as missing DNA sequence data for criti- Gray and McKinnon (2006) described in cal non-telmatherinid taxa (e.g. Craterocepha- detail mating behavior of seven telm- lus (Quirichtys) stramineus), suggests that ad- atherinid species from Lakes Matano (Telm- ditional data should be analyzed before fol- atherina antoniae, T. sarasinorum, T. wahjui, lowing this suggestion. T. “whitelips” (= T. “thicklip”) and To- Roy et al. (2004) investigated mtDNA wuti (Telmatherina celebensis, Tominanga sp., haplotype relationships of four fam- Paratherina sp.). They noted presence of ilies from L. Matano. They included the two distinct size classes in T. antoniae but Matano-endemics Oryzias matanensis for did not find differences in mating behav- Oryziidae, Dermogenys weberi for Hemiram- ior among these. Phenotypes referred here phidae, undetermined Glossogobius for Go- to Telmatherina opudi/T. obscura were treated biidae and undetermined Telmatherina for together with T. sarasinorum by Gray and Telmatherinidae. Together with findings McKinnon (2006). They also did not fol- from cursory morphometric analyses they low Kottelat’s concept of separating two conclude that the endemic diversity of L. species of Tominanga as they found Matano’s Telmatherina fulfills the criteria of discrepancies between the descriptions and an adaptive radiation. However, several parts own observations. However, according to in their argumentation cast serious doubts Kottelat, T. aurea is restricted to L. Maha- on their conclusions. First, discussion and lona which was not covered in that study. conclusions of this work focus on and rest Gray and McKinnon (2006) concluded that

144 mating behavior can be generalized in ing, with focus on L. Matano. Fishing tech- Malili Lakes telmatherinids, and they de- niques applied were SCUBA- and snorkel- scribed common intra- and intersexual ing-aided gillnetting, electro-fishing, dip- behavioral interactions. and beach-netting. At Lake Matano, sam- Herder et al. (2006) investigated phyloge- pling was focused on six major sampling netic relationships of Malili Lakes telm- locations distributed almost equally around atherinids, provided evidence for introgres- the lake (locations 1-6, fig. 2) and two addi- sive hybridization of stream phenotypes into tional sampling locations at the mouth of the most diverse clade of L. Matano’s telm- River Lawa (location 7) and the outlet to atherinids and concluded that hybridization River Petea (location 8). Each of the six could have acted as a key factor in the gen- major sampling locations was sampled by eration of the flock’s spectacular diversity. two or three investigators for five to seven They provided genetic (mitochondrial and days both in dry season 2002 and wet season multilocus nuclear DNA) and morpho- 2004. Voucher specimens are deposited in metric evidence supporting the assumption fish collections of ZSM (Zoologische of an adaptive radiation within L. Matano, Staatssammlung München), ZFMK (Zoo- with the original signal of monophyly logisches Forschungsmuseum Alexander masked secondarily by introgression. Her- Koenig, Bonn), and MZB (Museum Zoo- der et al. (2006) included a previously logicum Bogoriense, Cibinong). SL refers unknown diversity of riverine telmatheri- to “standard length”, measured from tip of nids in their analyses and found strong sup- snout to caudal flexure (Kottelat 1990a) port to divide them into three groups: (1) (= end of hypural plate). The terms “mor- Telmatherina sp. “Petea”, closely related to photypes” and “color morphs” are used to L. Matano’s “sharpfins”; (2) T. cf. bonti classify telmatherinid diversity in a prag- “West”, a separate, ancient clade; (3) all re- matic way as long as no detailed descrip- maining T. cf. bonti, including those from tions are available, without referring to spe- Rivers Nuha, Tominanga, Malili and streams cies concepts or claiming any kind of spe- entering Lakes Matano and Towuti. They cies status. Consequently, “morphotype” also included undescribed lake-dwelling refers to a group of phenotypically similar phenotypes from L. Matano, Telmatherina sp. individuals sharing consistent characters pres- “thicklip” ( = T. “whitelips” of Gray and ent in males and females. “Color morphs” McKinnon 2006); T. sp. “elongated”, and are analogously understood as different co- T. sarasinorum “bigmouth” and “largehead”, lor patterns expressed within “morphotypes”; phenotypes differing in the eponymous these are defined to males only, as no female characters from T. sarasinorum Kottelat. polychromatism was recorded. Moreover, they included T. cf. celebensis Color descriptions are based on field ob- “Lontoa”, a phenotype from L. Lontoa, servations directly after collection, with closely related to T. celebensis from L. Towu- focus to overall coloration of body and ti and Mahalona. fins. They describe patterns of general ap- pearance but do not provide detailed cha- 2. Material and methods racters for taxonomic purposes. The term “color morph” therefore represents a In surveys on telmatherinid diversity in dry pragmatic simplification and does not re- season 2002 and wet season 2004, all major flect the complete variance present in water bodies of the Malili Lake system were color characters. investigated by SCUBA and snorkeling ob- In Lake Matano, telmatherinids are separ- servations down to 20 m (lakes), and by fish- able in two groups of morphotypes, char-

Verh. Ges. Ichthyol. Bd. 5, 2006 145 Fig. 1: The Malili Lakes system is located in Sulawesi, Indonesia. Symbols represent sampling loca- tions of the surveys 2002 and 2004, with triangles for sampling locations at Lakes Towuti and Mahalona and black circles for locations at rivers and streams; numbers refer to table 1. © map by von Rintelen and Zitzler, modified (with permission). Abb. 1: Das Maliliseen- System befindet sich in Sulawesi, Indonesien. Symbole markieren Sammel- punkte aus den Untersuchungsjahren 2002 und 2004. Dreiecke stehen für Sammelstellen an den Seen Towuti und Mahalona, schwarze Kreise für Probestellen an Flüssen und Bächen; die Numme- rierung bezieht sich auf Tabelle 1. © der Karte liegt bei von Rintelen und Zitzler, modifiziert (mit freundlicher Erlaubnis).

Fig. 2: The telmatherinid fauna of Lake Matano was investigated in detail at six major sampling locations (1-6) as well as at the mouth of River Lawa entering the lake at its north-western corner (7) and at the outlet area draining the lake to River Petea (8). Record of telmatherinid phenotypes is provided in table 2. © map by von Rintelen and Zitzler, modified (with permission). Abb. 2: Die Telmatheriniden-Fauna des Matanosees wurde an sechs Hauptprobestellen (1-6) de- tailliert untersucht. Weitere Probestellen sind die Einmündung des Flusses Lawa am Nordwestende des Sees (7) sowie das „Outlet“-Gebiet, in dem der See in den Fluss Petea übergeht. Nachweis von Telmatheriniden- Phaenotypen findet sich in Tabelle 2. © der Karte liegt bei von Rintelen und Zitzler, modifiziert (mit freundlicher Erlaubnis).

146 Figs. 3a-j: “Morphotypic” and “color morph” diversity of “sharpfin” telmatherinids of Lake Matano. a Telmatherina abendanoni male, b Telmatherina abendanoni female, c Telmatherina opudi males, yellow and blue morph, d Telmatherina opudi male, yellow morph, e Telmatherina obscura male, blackish morph, f Telmatherina opudi female, g male, blue morph, h Telmatherina sarasinorum male, yellow morph, i Telmatherina sarasinorum male, grey morph, j Telmatherina sarasinorum female. Abb. 3a-j: Morphotypische und Farbformendiversität der „sharpfin“ Telmatheriniden des Mata- nosees. a Telmatherina abendanoni Männchen, b Telmatherina abendanoni Weibchen, c Telmatherina opudi Männchen, gelbe und blaue Morphe, d Telmatherina opudi Männchen, gelbe Morphe, e Telmatherina obscura Männchen, schwärzliche Morphe, f Telmatherina opudi Weibchen, g Telmatherina sarasinorum Männchen, blaue Morphe, h Telmatherina sarasinorum Männchen, gelbe Morphe, i Telmatherina sarasi- norum Männchen, graue Morphe, j Telmatherina sarasinorum Weibchen.

Verh. Ges. Ichthyol. Bd. 5, 2006 147 acterized by shape of their second dorsal 3.1. Telmatherina of Lake Matano and anal fins, either pointed (“sharpfins”) or rounded (“roundfins”) (Herder et al. Seven “morphotypes” of “sharpfins” (figs. 2006, Kottelat 1991). To account for the 3a-j, 4a-h) and three of “roundfins” (figs. striking size dimorphism in Telmatherina 5a-j, 6a) are distinguished within L. Matano antoniae, “small” and “large” phenotypes (tabs. 1 and 2). Five “sharpfin” morphotypes are differentiated according to a gap in are di- or polychromatic in male coloration size frequency distribution of adult speci- (tab. 1; figs. 3c-e, g-i and 4a-b, d-e). Three mens between 55 and 60 mm SL. previously unknown “sharpfin” phenotypes are recognized and can be identified pre- 3. Results liminarily: Telmatherina sp. “thicklip” (= T. “whitelips” in Gray and McKinnon 2006), All telmatherinid populations observed are characterized by pronounced fleshy lips and sexually dimorphic. “Morphotypes” and extremely prolonged tips of second dorsal male “color morphs” are summarized in and anal fins (fig. 4g-h); T. sp. “elongated”, table 1 and illustrated in figures 3 to 9; sample with short fins and an extremely slender sites are marked in figures 1 and 2. Lake body (figs. 4d-f); and T. sarasinorum “large- Masapi was sampled intensively at two head / bigmouth”, with a conspicuously sites, without recording any telmatherinid large head / big mouth (figs. 4a-b). Telma- specimen. therina abendanoni is recognized in accordance

Tab. 1: List of telmatherinid species/“morphotypes“ inhabiting the Malili Lakes system and sur- rounding catchments, their color morphs and distribution within the Malili Lakes system. • = new record, • = previous record confirmed, * = only juveniles obtained, ** = not confirmed Tab. 1: Auflistung der nachgewiesenen Telmatheriniden-Arten/„Morphotypen“ des Maliliseen- Systems und umgebender Einzugsgebiete, ihrer Farbmorphen und Verbreitung innerhalb des Ma- liliseen-Systems. • = neuer Nachweis, • = vorheriger Nachweis bestätigt; * = nur Jungfische gefan- gen, ** = nicht bestätigt

148 Tab. 2: Record of telmatherinid “morphotypes” within Lake Matano. “Clade” refers to “sharpfins” and “roundfins” which have been identified as two major clades in Lake Matano according to nuclear DNA markers (Herder et al. 2006). Tab. 2: Nachweis von Telmatheriniden- „Morphotypen“ innerhalb des Matanosees. “Clade” be- zieht sich auf “sharpfins” und “roundfins”, die mit DNA-Kernmarkern als zwei Hauptclades in- nerhalb des Matanosees identifiziert wurden (Herder et al. 2006).

to Kottelat’s (1991) concept of the species mm SL) and a slender body (figs. 5e-g), in as large and slender, predatory “sharpfin” contrast to larger size (>60 mm SL) and (figs. 3a-b). It appears together with Telma- deeper body in T. antoniae “large” (figs. 5a-d) therina sp. “thicklip” the only of L. Matano’s (Kottelat 1991, Herder et al., unpubl.). Telm- telmatherinids without color polymor- atherina prognatha, a large predatory pheno- phism. Telmatherina opudi and T. obscura are type, is as large as T. antoniae “large”, but as characterized by only moderately elonga- slender as T. antoniae “small”, and is charac- ted fins and in having a comparatively steep- terized by having conspicuous “beak-like” ly angled mouth. Following the concept of jaws (figs. 5h-j and 6a) (Kottelat 1991). In Kottelat (1991), the blackish color morph all three “roundfins”, males occur in three is called T. obscura (figs. 3e), the yellow and color variants: yellow, blue or blue-yellow bluish-grey T. opudi (figs. 3c-d, f). Telmathe- (bluish body and yellow fins). rina sarasinorum is recognized according to Whereas male roundfins” and “sharpfins” Kottelat (1991) by its small to medium-si- can be distinguished according to the shape zed mouth and conspicuous elongated se- of their second dorsal and anal fins (Kot- cond dorsal and anal fins in males. Bright telat 1991), females of both groups are sep- yellowish, blackish-blue and blackish-grey arable according to body coloration which color morphs are present (Kottelat 1991; is silver-grey in “roundfins” (figs. 5d, g and figs. 3g-i). Telmatherina wahjui was described 6a) vs. yellowish dusky-brown in “sharp- by Kottelat (1991) as wide-headed “sharp- fins” (figs. 3f, j and 4f, h); both can exhibit fin” endemic to the outlet region, the small bluish reflections. Females can like males transition area between L. Matano and R. be assorted to the “morphotypes” listed Petea (fig. 4c). Its diagnostic character is above. relation of snout width to snout length, re- sulting in a blunt, wide snout in contrast to 3.2. Telmatherina of Lakes Towuti, the more elongated snout in T. opudi and T. Mahalona and Lontoa sarasinorum. There are two male color morphs in T. wahjui, either blue or yellow. Telmatherina celebensis was by far the most Lake Matano’s “roundfins” are distinguish- abundant telmatherinid observed in Lakes able according to two distinct size classes Towuti and Mahalona. Male color morphs of adults in Telmatherina antoniae: T. antoniae with either blue or yellow second dorsal “small” is characterized by small size (<55 and anal fins were observed frequently,

Verh. Ges. Ichthyol. Bd. 5, 2006 149 Figs. 4a-h: “Morphotypic” and “color morph” diversity of “sharpfin” telmatherinids of Lake Matano. a Telmatherina sarasinorum “bigmouth” male, yellow morph, b Telmatherina sarasinorum “large- head” male, blackish morph, c Telmatherina wahjui male, blue morph, d Telmatherina sp. “elongated” male, yellow morph, e Telmatherina sp. “elongated” male, blue morph, f Telmatherina sp. “elongated” female, g Telmatherina sp. “thicklip” male, h Telmatherina sp. “thicklip” female. Abb. 4a-h: Morphotypische und Farbformendiversität der „sharpfin“ Telmatheriniden des Mata- nosees. a Telmatherina sarasinorum “Großmaul” Männchen, gelbe Morphe, b Telmatherina sarasinorum “Großkopf ” Männchen, schwärzliche Morphe, c Telmatherina wahjui Männchen, blaue Morphe, d Telmatherina sp. “verlängert” Männchen, gelbe Morphe, e Telmatherina sp. “verlängert” Männchen, blaue Morphe, f Telmatherina sp. “verlängert” Weibchen, g Telmatherina sp. “Dicklippe” Männchen, h Telmatherina sp. “Dicklippe” Weibchen.

150 Figs. 5a-j: “Morphotypic” and “color morph” diversity of “sharpfin” telmatherinids of Lake Matano (continued). a Telmatherina antoniae “large” male, blue morph, b Telmatherina antoniae “large” male, yel- low morph, c Telmatherina antoniae “large” male, blue-yellow morph, d Telmatherina antoniae “large” female, e Telmatherina antoniae “small” males, blue and yellow morph, f Telmatherina antoniae “small” male, blue-yellow morph, g Telmatherina antoniae “small”, female, h Telmatherina prognatha male, blue morph, i Telmatherina prognatha male, yellow morph, j Telmatherina prognatha male, blue-yellow morph. Abb. 5a-j: Morphotypische und Farbformendiversität der „sharpfin“ Telmatheriniden des Mata- nosees (Fortsetzung). a Telmatherina antoniae „groß” Männchen, blaue Morphe, b Telmatherina antoniae „groß” Männchen, gelbe Morphe, c Telmatherina antoniae „groß” Männchen, blaugelbe Morphe, d Telmatherina antoniae „groß” Weibchen, e Telmatherina antoniae „klein” Männchen, blaue und gelbe Morphe, f Telmatherina antoniae „klein” Männchen, blaugelbe Morphe, g Telmatherina antoniae „klein”, Weibchen, h Telmatherina prognatha Männchen, blaue Morphe, i Telmatherina prognatha Männchen, gelbe Morphe, j Telmatherina prognatha Männchen, blaugelbe Morphe.

Verh. Ges. Ichthyol. Bd. 5, 2006 151 however both being characterized by a pre- 3.4. Paratherina dominantly yellow to yellowish-brown body (figs. 6b-c). Telmatherina celebensis was Paratherina sp. comprise the largest pheno- also recorded within R. Tominanga. In types of Telmatherinidae. The largest is Lake Lontoa, Telmatherina cf. celebensis, a (largest specimen obtained: phenotype resembling T. celebensis, but dis- 142.6 cm SL), characterized by its slightly tinguished by its conspicuous deep body, superior or isognathous mouth, bluish body is present (figs. 6d-f). Like T. celebensis, it is also and conspicuous faint lateral stripes over a characterized by yellow-blue male dichroma- “fishbone-like” transversal pattern (figs. 7e- tism (figs. 6d-e). Female specimens of both f, j). Two male colorations were observed, phenotypes are dusky yellowish brown with having either bluish white (fig. 7f) or yel- hyaline fins. Telmatherina cf. celebensis “Lontoa” low (fig. 7e) margins of second dorsal, anal was the single telmatherinid species obtained and caudal fins. Paratherina cf. cyanea is a very from L. Lontoa in our surveys. slender, bluish Paratherina with conspicu- ously large eyes (fig. 8a). However, our ma- 3.3. Tominanga terial of this species is restricted to few ob- viously immature specimens. Paratherina cf. The genus Tominanga exhibits the most wolterecki (figs. 7h, i) is characterized in hav- complex pattern of male colorations in ing a strongly upwards directed, superior telmatherinids, with varying degrees of red, mouth. Males possess blackish yellow, slightly yellow, silver and blackish patterns (e.g. figs. elongated fins, in contrast to females char- 6g-j and 7a, c-d). In contrast, females are acterized by short, pale yellowish to hyaline uniformly colored, with silvery body and fins. hyaline fins (fig. 7b). Patterns of male col- oration are not restricted to those described 3.5. Riverine phenotypes – Telmathe- by Kottelat (1990a), who distinguished rina cf. bonti color patterns of T. sanguicauda inhabiting L. Towuti with blue or red body, red cau- Telmatherinids were found to be present in dal fin, blue or red dorsals, anal and pelvic several rivers and streams, including Rivers from T. aurea restricted to L. Mahalona, Petea (figs. 8b-c, 9f) and Tominanga (figs. diagnosed by silvery body with yellow or 8g, 9g), and in some rivers outside of the red spots, caudal, dorsals and pelvic in plain Malili Lakes watershed (fig. 1, tab. 1). River- red or yellow or hyaline with yellow or red ine telmatherinids exhibit a diversity of dots. Illustrations presented here display body- and fin-shapes, ranging from slender only a fraction of color variation observed (e.g. figs. 8f and 9a) to rather high-bodied in male Tominanga in the field, including (e.g. figs. 8b-c, g, j), and rounded (e.g. fig. plain yellow to golden specimens in L. 8e) to pointed (e.g. figs. 8b-d, f-h, j) fins. Mahalona (fig. 6h), individuals with silvery Corresponding to the original description to yellowish body and various degrees of of Telmatherina bonti and consistent to yellow sparkles (figs. 6g, i), bluish body Kottelat’s descriptions (Kottelat 1990a), with red fins (fig. 6j) or bluish body with phenotypes from Lakes Mahalona and To- red spots in second dorsal and anal fin wuti, as well as from streams draining into combined with orange caudal (fig. 7a). L. Towuti and to some extent also those Besides from Lakes Towuti and Mahalo- from River Lawa exhibit a conspicuous na, Tominanga are here also reported to “hump” at the nape (e.g. figs. 8h-j). In con- occur within River Tominanga and the trast, those from River Lemulengku (fig. 8e), lower stretches of R. Petea. R. Saluro (fig. 8f), isolated Rivers Nuha (figs.

152 9b-d) and “West” (fig. 9a), are more slen- atherinid populations in a landmark-based der and lack the “hump”. Phenotypes ob- morphometric analysis of body shape. They tained from R. Tominanga show a conspi- demonstrated that there is low overlap in cuous sub-terminal mouth (fig. 8g); like tho- morphospace among “stream clade” telm- se from R. Petea, they are characterized by atherinids and the adaptive radiation of L. a rather high body. Females of all riverine Matano’s “roundfins” and “sharpfins”, populations are characterizes by brownish- though “sharpfins” are heavily introgressed grey body and hyaline, short fins (e.g. figs. by stream telmatherinids. The same analy- 8i and 9d). Color polymorphisms in males sis also provided first empirical hints sug- are present in several populations, ranging gesting inter-population variation in mor- from white/orange-red in River Nuha (figs. phometric traits. Forthcoming analyses 9b-c) and yellow/bluish in R. Petea (figs. (Herder and Schliewen, in prep.) will evalu- 8b, c) and R. “Beau” to yellow/white in R. ate interrelationship of this strongly unde- “West” (fig. 9a). The dominating fin colo- restimated group using population genetic rations of non-polymorphoc riverine popu- and morphometric tools. However, indepen- lations are orange-red, yellow or white (figs. dent of species status for different riverine 8d-h, j; tab.1). populations it appears necessary to point In accordance to previous records (Kottelat out the high levels of phenotypic and ge- 1990a, Weber and Beaufort 1922), Telma- notypic diversity of telmatherinids inhabi- therina bonti was in the present surveys ob- ting different flowing waters of the Malili tained directly from L. Towuti. We include Lakes system, because this fact highlights here also a record from Lake Mahalona clo- the importance of fine-scaled watershed se to the mouth of R. Petea. management within the larger Malili area including damming and mining plans in se- 4. Discussion lected sub-watersheds.

Exceeding Kottelat’s revision (Kottelat 4.2. Telmatherina of Lake Matano 1990a, 1991) listing 15 described telmatheri- nid species within the Malili Lakes system, All telmatherinid phenotypes of Lake Mata- five of which are polymorphic in male co- no are endemic to this lake. Complexity of loration, we presently distinguish 19 “mor- Lake Matano’s adaptive sailfin-silverside photypes” from lake (incl. T. cf. bonti from radiation is higher than initially expected, Lakes Towuti/Mahalona) and additional 13 but appears still manageable for approa- from riverine habitats. In total, 17 of these ching questions regarding the analysis of exhibit di- or polychromatic male ornamen- the flock’s genesis. There are no hints for tation (tab. 1). clear allopatric restrictions of telmatheri- nids within the lake, apart from T. wahjui 4.1. The diversity of stream phenotypes that is described from the outlet area (tab. 2; Kottelat 1991). Lake Matano’s outlet area Most of the additional “morphotypes” (13) at the mouth of River Petea is here under- originate from riverine habitats like rivers stood as an approx. 100 meters long transi- and streams. Previous studies did not diffe- tion zone between the lake and the stream. rentiate riverine telmatherinids further and This area is characterized by medium to referred to all phenotypes as Telmatherina strong current, coarse substrate consisting bonti. A first step to analyzing this diversi- mainly of empty bivalve shells and low ty was conducted by Herder at al. (2006), depth of up to 1.5 m. These characteristics who included nine stream-dwelling telm- might explain the absence of most lake in-

Verh. Ges. Ichthyol. Bd. 5, 2006 153 Figa. 6a-j: “Morphotypic” and “color morph” diversity of “roundfin” telmatherinids of Lake Matano. a Telmatherina prognatha female, b Telmatherina celebensis male, yellow-blue morph (Mahalona), c Telmatherina celebensis male and female (Towuti), d Telmatherina cf. celebensis “Lontoa” male, yellow morph, e Telmatherina cf. celebensis “Lontoa” male, blue morph, f Telmatherina cf. celebensis “Lontoa” female, g Tominanga cf. aurea male, silver-yellow morph (Mahalona), h Tominanga cf. aurea male, yel- low morph (Mahalona), i Tominanga cf. aurea male, silver-sparkle-yellow morph (Mahalona), j Tomin- anga cf. aurea male, silver-red morph (Mahalona). Abb. 6a-j: Morphotypische und Farbformendiversität der „roundfin“ Telmatheriniden des Matano- sees. a Telmatherina prognatha Weibchen, b Telmatherina celebensis Männchen, gelbblaue Morphe (Mahalo- na), c Telmatherina celebensis Männchen und Weibchen (Towuti), d Telmatherina cf. celebensis „Lontoa” Männchen, gelbe Morphe, e Telmatherina cf. celebensis „Lontoa” Männchen, blaue Morphe, f Telmatheri- na cf. celebensis „Lontoa” Weibchen, g Tominanga cf. aurea Männchen, silbergelbe Morphe (Mahalona), h Tominanga cf. aurea Männchen, gelbe Morphe (Mahalona), i Tominanga cf. aurea Männchen, silbern ge- sprenkelte gelbe Morphe (Mahalona), j Tominanga cf. aurea Männchen, silberrote Morphe (Mahalona).

154 Figs. 7a-j: “Morphotypic” and “color morph” diversity of Tominanga (a-d) and Paratherina (e-j). a Tominanga cf. aurea male, blue-red morph (Mahalona), b Tominanga sp. female (Mahalona), c Tominan- ga cf. sanguicauda male, silver-red-black morph (Towuti), d Tominanga cf. sanguicauda male, silver- orange morph (Towuti), e Paratherina striata male, yellow morph (Mahalona), f Paratherina striata male, blue morph (Mahalona), g Paratherina striata female (Mahalona), h Paratherina cf. wolterecki male (Mahalona), i Paratherina cf. wolterecki female (Mahalona), j Paratherina striata female (Towuti). Abb. 7a-j: Morphotypische und Farbformendiversität von Tominanga (a-d) and Paratherina (e-j). a Tominanga cf. aurea Männchen, blaurote Morphe (Mahalona), b Tominanga sp. Weibchen (Mahalona), c Tominanga cf. sanguicauda Männchen, silberrotschwarze Morphe (Towuti), d Tominanga cf. sanguicau- da Männchen, silberorange Morphe (Towuti), e Paratherina striata Männchen, gelbe Morphe (Maha- lona), f Paratherina striata Männchen, blaue Morphe (Mahalona), g Paratherina striata Weibchen (Ma- halona), h Paratherina cf. wolterecki Männchen (Mahalona), i Paratherina cf. wolterecki Weibchen (Maha- lona), j Paratherina striata Weibchen (Towuti).

Verh. Ges. Ichthyol. Bd. 5, 2006 155 habiting telmatherinids except for T. wahjui 4.3. Telmatherina of Lakes Towuti and T. opudi. However, a few specimens and Mahalona identifiable to Kottelat’s description of T. wahjui (Kottelat 1991) were also obtained at The presence of Telmatherina celebensis in other locations within the lake, all of them River Tominanga is the first record of the in the extreme shallow waters of the surge species in a riverine habitat, which might area (tab. 2). point to ongoing gene flow between popu- Lake Matano’s “sharpfins” are conspicu- lations of lakes Towuti and Mahalona. Dis- ously more diverse than the “roundfins”, tinctiveness of Telmatherina cf. celebensis from with regard to (1) the number of “morpho- Lake Lontoa with respect to T. celebensis from types” and to (2) character diversity within the large lakes is currently being investigated “morphotypes” (for an analysis of mor- (Hadiaty and Herder, in prep). phospace, see Herder et al. 2006). Diver- sity of “sharpfins” trophic morphology with 4.4. Paratherina respect to trophic ecology suggests that the process of divergence and speciation is A number of presently unidentifiable speci- ongoing, a hypothesis which is presently mens as well as field observations of tested (Pfaender et al., in prep.). In this con- unknown phenotypes strongly suggest that text, further analyses are required for un- species number of Paratherina is currently derstanding patterns of morphological di- underestimated and that Paratherina is only versity in T. sarasinorum/T. sarasinorum insufficiently covered by our survey. How- “largehead/bigmouth” and in T. opudi. ever, additional material is available in the Kottelat (1991) described male color dimor- fish collection of MZB and is presently used phisms in two of five “sharpfin” species. to revise the genus (Hadiaty and Herder, in Here, we add records of Telmatherina opudi prep). Therefore, the present determinations “yellow”, T. sarasinorum “blackish”, T. sp. are given with provisio. “elongated yellow”, T. sp. “elongated blue” Paratherina striata was previously only re- and T. sp. “thicklip“. corded from Lakes Towuti and Lontoa, We were not able to distinguish Telmathe- though Kottelat (1990a) assumed it to be rina opudi and T. obscura following Kottelat’s present in L. Mahalona, too. Here, we pro- (1991) diagnoses using other characters than vide the first record for L. Mahalona. The coloration. However, due to our observa- species is dimorphic in male coloration, with tions, intermediate phenotypes among T. sa- either yellow or blue margins of second rasinorum, T. opudi and T. obscura exist, and dorsal and anal fins. For Paratherina cf. cya- phenotypic diversity present in “sharpfins” nea, present sampling is limited to few speci- is striking. Detailed morphological and mens from L. Towuti; according to Kot- molecular analyses of large series are re- telat (1990a), it is also present in L. Maha- quired to understand the evolutionary pro- lona. Upcoming analysis of morphological cesses shaping diversity of this group and characters incorporating Kottelat’s material to deduce taxonomic conclusions. is aimed to clarify characters and infer the For “roundfins“, we report apart from yel- species’ distribution. Paratherina cf. wolterecki low and blue male morphs in Telmatherina has been reported by Kottelat (1990a) from antoniae “large” and T. antoniae “small” the Lakes Mahalona and Towuti. Here, we only occurrence of blue-yellow phenotypes hav- confirm the species for its type locality, L. ing a bluish body and yellow fins. For T. Mahalona. No male color dimorphism was prognatha, we add “yellow” and “blue-yel- recorded in this species. Paratherina labiosa low” male morphs to the known blue one. has so far not been identified from the pre-

156 sent material according to the available keys ing an original origin from River Tominan- (Aurich 1935, Kottelat 1990a). ga. The only published previous record of Neither the presence of P. striata nor P. la- the genus in flowing water is that of T. san- biosa, both recorded by Aurich (1935) for guicauda in its original description (Kottelat Lake Lontoa, were confirmed. However, the 1990a), reported to occur in a cold tribu- present sampling is restricted to the results tary of L. Towuti. of intensive fishing and snorkeling at a single site at the south-western corner of 4.6. Threats to the lake’s aquatic the lake. A denser sampling scheme is re- diversity quired for judging about Paratherina pres- ence in L. Lontoa. Around this lake, the Over the last decades, nickel mining by PT environment has undergone substantial de- INCO became a major industry at Soroako gradation by deforestation, with almost near L. Matano. Substantial parts of the area complete absence of forest over large areas. south of L. Matano have already been As descriptions of the original situation are cleared and opened to surface mining, areas lacking, the presently observable low trans- east and south-east of the lake and nor- parency of the lake’s waters can only specu- thwards of R. Petea are presently under lated to be a result of deforestation. Ab- digging. Exploited nickel-ore is smelted on sence of Paratherina may be a result of those site in several furnaces, causing environ- substantial environmental changes, too. mental pollution and requiring gigantic amounts of electric energy. Hydroelectric 4.5. Tominanga power plants driven by a series of dams along R. Larona are used to produce the In Tominanga, red, orange, yellow and necessary electricity. However, substantial blackish patterns dominate male color poly- efforts are undertaken by INCO to restrict morphism. Male Tominanga apparently re- environmental impact on the lakes, es- present the chromatically most diverse pecially on L. Matano. According to Gie- group of telmatherinids: Different pheno- sen et al. (1991), pollution via atmospheric types exhibit a mosaic of color components deposition (SO2 dissolved in rainwater) is with plain or speckled patterns of yellow, unlikely to lower the lake’s pH due to high red, orange, blackish and/or silvery colors. buffering capacity of its water. In contrast, In contrast, no external morphological dif- contaminated effluents from the plant do ferentiation is visible, neither among speci- represent a serious threat, measurable as mens within each of both lakes nor among increased levels of sulphur in lake and color morphs within these. However, de- stream waters (Giesen et al. 1991). tailed morphological examinations based on The major threats to the endemic aquatic representative series of all Tominanga color fauna are represented undoubtedly by morphs are lacking and might reveal indi- INCO’s activities for water regulation besi- cations supporting species status of Kotte- des erosion by increased land-use and logg- lat’s (1990a) T. aurea and T. sanguicauda. Un- ing. River Larona has already been trans- til this question is solved, we suggest using formed from a natural river to an artificial “Tominanga cf. aurea” for specimens from channel, L. Towuti and R. Larona are L. Mahalona, and “Tominanga cf. sanguicau- dammed, and mining with massive earth- da” for those from L. Towuti. moving goes on along R. Petea (Herder, The present work provides the first record pers. obs.). Giesen et al. (1991) list additional of the genus Tominanga in Rivers Petea and plans of INCO to manage water resources Tominanga, despite its etymology suggest- of the Malili Lakes system, including ca-

Verh. Ges. Ichthyol. Bd. 5, 2006 157 Figs. 8a-j: “Morphotypic” and “color morph” diversity of Paratherina (a) and riverine phenotypes Telmatherina cf. bonti (b-j). a Paratherina cf. cyanea (Towuti), b Telmatherina cf. bonti “Petea” male, yellow morph, c Telmatherina cf. bonti “Petea” male, blue morph, d Telmatherina cf. bonti “Lawa” male, yel- low-red morph, e Telmatherina cf. bonti “Lemulengku” males, f Telmatherina cf. bonti “Saluro” males, g Telmatherina cf. bonti “Tominanga” male, h Telmatherina cf. bonti “South Towuti” male, i Telmatherina cf. bonti “South Towuti” female, j Telmatherina cf. bonti “Lemolemo” male. Abb. 8a-j: Morphotypische und Farbformendiversität von Paratherina (a) und flussbewohnenden Phaenotypen Telmatherina cf. bonti (b-j). a Paratherina cf. cyanea (Towuti), b Telmatherina cf. bonti „Pe- tea” Männchen, gelbe Morphe, c Telmatherina cf. bonti „Petea” Männchen, blaue Morphe, d Telmathe- rina cf. bonti „Lawa” Männchen, gelbrote Morphe, e Telmatherina cf. bonti „Lemulengku” Männchen, f Telmatherina cf. bonti „Saluro” Männchen, g Telmatherina cf. bonti „Tominanga” Männchen, h Telm- atherina cf. bonti „South Towuti” Männchen, i Telmatherina cf. bonti „South Towuti” Weibchen, j Telm- atherina cf. bonti „Lemolemo” Männchen.

158 Figs. 9a-g: “Morphotypic” and “color morph” diversity of riverine phenotypes Telmatherina cf. bonti (a-d) and habitat illustrations of the rivers interconnecting or draining the three major lakes (e-g). a Telmatherina cf. bonti “West” males, blue and yellow morph, b Telmatherina cf. bonti “Nuha” male, orange morph, c Telmatherina cf. bonti “Nuha” male, white morph, d Telmatherina cf. bonti “Nuha” female, e River Larona, f River Petea, g River Tominanga. Abb. 9a-g: Morphotypische und Farbformendiversität flussbewohnender Phänotypen Telmatherina cf. bonti (a-d) und Abbildung der Flüsse, die die drei Hauptseen verbinden bzw. entwässern (e-g). a Telmatherina cf. bonti „West” Männchen, blaue und gelbe Morphe, b Telmatherina cf. bonti „Nuha” Männchen, orange Morphe, c Telmatherina cf. bonti „Nuha” Männchen, weiße Morphe, d Telmatheri- na cf. bonti „Nuha” Männchen, e Laronafluss, f Peteafluss, g Tominangafluss.

Verh. Ges. Ichthyol. Bd. 5, 2006 159 nalization of Rivers Petea and Tominanga, undetected diversity of stream telmatheri- construction of a channel leading waters nids and announced to preserve Rivers Pe- from L. Masapi to L. Towuti, redirection tea and Tominanga by establishing buffer of the northwards-draining River Nuha into areas between mining areas and the rivers, L. Matano, erection of an additional dam accompanied by abandonment of canaliza- at R. Larona (already reality), construc- tion. tion of a sluice on R. Petea to store water Finally, introduction of foreign fish species in L. Matano and permanent increase of is another potential threat for the lakes en- water level in L. Towuti by damming. demic fauna, a topic under investigation in These activities may have extreme impacts context of our recent surveys (Herder et on the endemic aquatic diversity of the al. in prep.). Introduction of not native fish Malili-Lake System and should be ad- species can cause serious threats to the na- justed to sustainable preserve this unique tive fauna (Kottelat et al. 1993, Whitten et lake system. al. 2002) and has already been reported to Additional threats come from habitat loss affect endemic radiations in Sulawesi (Kot- due to mining activities and population in- telat 1990d). Therefore, we strongly demand crease. For example, streams in the former to avoid any introduction of not mining area south-east of L. Matano have native to one of the lakes ecosystems. been found either not to exist anymore or to have changed to biologically dead re- 5. Conclusions mains, filled with mud, oil and waste; only one of these contained fish, which, how- Telmatherinidae of the Malili Lakes are ever, were stocked African “Tilapia” (Her- much more diverse than previously known, der, pers. obs.). Likewise, streams north-east with regard to both, “morphotypic” diver- of L. Matano have been found to be ex- sity and color morphs. Male color di- or tremely degraded by oil contamination, mass- polymorphism is the striking characteristic ive gathering of water and input of waste. of this species-flock and is found in the However, this area is an upcoming mining area majority of species/“morphotypes”, includ- and will also be lost within near future. ing all three presently recognized genera. A recent meeting on “The Ecology and Among lacustrine telmatherinids, diversity Limnology of the Malili Lakes” (Bogor, is highest in L. Matano’s “sharpfins”. With Indonesia 20-22 March 2006) organized by regard to lacustrine species of Lakes To- PT INCO initiated discussion and intellec- wuti and Mahalona, a close faunistic rela- tual exchange among scientists from various tionship between both water bodies is evi- fields, all conducting research at the Malili dent. An initially unknown diversity of stream- Lakes. It became obvious that there exists a dwelling telmatherinids was discovered and deficiency among scientific knowledge, remains to be studied in detail. mostly scattered to specialists, and informa- The adaptive radiation of sailfin-silversides tion received and finally taken into consider- has great potential to become a reference ation by PT INCO and the local author- system allowing evolutionary biologists to ities. The leading representative of PT test ideas regarding radiation processes and INCO expressed that his company has deep maintenance of color polymorphisms, using interest to minimize environmental impacts a compact model system of manageable to the lakes and assured that scientific re- diversity. Comparative approaches for in- sults available on the lake system’s ecology vertebrate radiations bear the chance to test will be considered in upcoming planning. the influence of various parameters to dif- In this context, he recognized the previously ferent groups of organisms. Probably the

160 most interesting point for speciation re- by T. von Rintelen. We are grateful to J. search is represented by the various degrees McKinnon and S. Gray for inspiring dis- of isolation between the lakes, allowing to cussions on telmatherinid diversity. The approach the impact of introgressive hy- manuscript benefited from critical com- bridization to adaptive radiations under sym- ments and constructive suggestions by T. patric, parapatric and allopatric scenarios. von Rintelen and A. Nolte. M. Kottelat However, present and planned construc- shared knowledge on telmatherinids and tions aimed to intensify usage of the lakes provided loan material. W. Wickler kindly water resources for energy supply as well supported a pilot study, which enabled suc- as massive loss of stream habitats due to cessful grant application. This study was mining activities are a major threat to the funded by research grants from the Deut- aquatic diversity. To avoid that the ende- sche Forschungsgemeinschaft to U.K.S. and mic radiations of the Malili Lakes system by a graduate fellowship donated by the become natural history in near future, pol- Rheinische Friedrich Wilhelms-Universität icy of water management has to take into Bonn to F.H. account the biodiversity of streams and rivers and natural distribution patterns of Literature the organisms within the lakes system. Drainages which are separated naturally Aarn, Ivantsoff, W. 1997. Descriptive anatomy should not inconsiderately be connected by of Cairnsichthys rhombosomoides and Iriatherina man, and the rivers between the lakes should werneri (Teleostei: Atheriniformes), and a phylo- be treated as vulnerable pathways. Also, genetic analysis of Melanotaeniidae. Ichthyol. complete destruction of stream drainages Explor. Freshwaters 8, 107-150. within current and upcoming mining areas Aarn, Ivantsoff, W., Kottelat, M. 1998. Phyloge- netic analysis of Telmatherinidae (Teleostei: should be avoided to allow, at least partial, Atherinomorpha), with description of Maros- for recolonization of ecosystems after ex- atherina, a new genus from Sulawesi. Ichthyol. ploitation of nickel ore. The authors are Explor. Freshwaters 9, 311-323. convinced that long-term sustainability of Abendanon, E.C. 1915a. Midden-Celebes-Expe- mining around the Malili Lakes can be in- ditie. Geologische en geographische Door- creased significantly by incorporating the kruising van Midden-Celebes (1909-1910), available and upcoming scientific results Atlas. E. J. Brill, Leiden. into management decisions. Abendanon, E.C. 1915b. Midden-Celebes-Expe- ditie. Geologische en geographische Door- kruising van Midden-Celebes (1909-1910), Acknowledgements vol. 2. E. J. Brill, Leiden. Ahl, E. 1936. Beschreibung eines neuen Fisches We thank LIPI for the permit to conduct der Familie Atherinidae aus Celebes. Zool. research in Indonesia. We are deeply grate- Anz. 114, 175-177. ful to INCO for their outstanding logistic Ahmad, W. 1977. Geology along the Matano Fault support at the lakes. F.H. acknowledges Zone East Sulawesi, Indonesia. Regional Con- INCO for sponsoring the symposium on ference on the Geology and Mineral resources “The Ecology and Limnonogy of the Ma- of South East Asia, Jakarta. lili Lakes” and thanks W. Napier and L. Aurich, H.J. 1935. Mitteilungen der Wallacea- Expedition Woltereck. Mitteilung XIV. Fische Hulet (INCO) for open-minded discussions II. Fam. Atherinidae. Zool. Anz. 112, 161-177. on conservation issues. For invaluable as- Boulenger, G.A. 1897. An account of the fresh- sistance in the field we thank J. Herder, A. water fishes collected in Celebes by Drs P and Nolte and J. Frommen. Fieldwork greatly F. Sarasin. Proc. Zool. Soc. Lond., Spec. Proc. benefited from logistic support in Indonesia Sarasin Expedition: 426-429.

Verh. Ges. Ichthyol. Bd. 5, 2006 161 Bramburger, A.J., Haffner, G.D., Hamilton, P.B. Kottelat, M. 1989a. Der Matano-See. DATZ 42, 2004. Examining the distributional patterns 616-618. of the diatom flora of the Malili Lakes, Sula- Kottelat, M. 1989b. Der Towuti-See. DATZ 42, wesi, Indonesia. Proc. 17th Int. Diatom Symp. 681-684. Ottawa, Canada, 11-25. Kottelat, M. 1989c. Die Süsswasser-Fauna von Bramburger, A.J., Haffner, G.D., Hamilton, P.B., Sulawesi. DATZ 42, 555-558. Hinz, F., Hehanussa, P.E. 2006. An examina- Kottelat, M. 1990a. Sailfin silversides (Pisces: tion of species within the genus Surirella from Telmatherinidae) of Lakes Towuti, Mahalona the Malili Lakes, Sulawesi Island, Indonesia, and Wawontoa (Sulawesi, Indonesia) with de- with descriptions of 11 new taxa. Diatom Res. scriptions of two new genera and two new spe- 21, 1-56. cies. Ichthyol. Explor. Freshwaters 1, 35-54. Brooks, J.L. 1950. Speciation in ancient lakes Kottelat, M. 1990b. The ricefishes (Oryziidae) (concluded). Quart. Rev. Biol. 25, 131-176. of the Malili Lakes, Sulawesi, Indonesia, with Crowley, L.E.L.M., Ivantsoff, W. 1992. Redefi- description of a new species. Ichthyol. Ex- nition of the freshwater fish genus Crateroce- plor. Freshwaters 1, 151-166. phalus (Teleostei: Atherinidae) of Australia and Kottelat, M. 1990c. Der Mahalona-See. DATZ New Guinea with an analysis of three spe- 43, 485-488. cies. Ichthyol. Explor. Freshwaters 3, 273-287. Kottelat, M. 1990d. Synopsis of the endangered Dyer, B.S., Chernoff, B. 1996. Phylogenetic rela- Buntingi (Osteichthyes: Adrianichthyidae and tionships among atheriniform fishes (Teleos- Oryziidae) of Lake Poso, Central Sulawesi, tei: Atherinomorpha). Zool. J. Linn. Soc. 117, Indonesia, with a new reproductive guild and 1-69. descriptions of three new species. Ichthyol. Giesen, W., Baltzer, M., Baruadi, R. 1991. Inte- Explor. Freshwaters 1, 49-67. grating conservation with land-use develop- Kottelat, M. 1991. Sailfin silversides (Pisces: ment in wetlands of South Sulawesi. Direc- Telmatherinidae) of Lake Matano, Sulawesi, torate General of Forest Protection and Na- Indonesia, with descriptions of six new spe- ture Conservation, Bogor. cies. Ichthyol. Explor. Freshwaters 1, 321-344. Gray, S., McKinnon, J.S.. 2006. A comparative Kottelat, M. 1992. Neu beschriebene Fische von description of mating behavior in the ende- Sulawesi. DATZ 45, 653-659. mic telmatherinid fishes of Sulawesi’s Malili Kottelat, M., Suttner, E. 1988. Catalogue des ty- Lakes. Env.. Biol. Fishes 75, 471-482. pes de poissons du Musée d’histoire naturel- Hadiaty, R.K., Wirjoatmodjo, S. 2002. Studi le de Bâle (Naturhistorisches Museum Basel). pendahuluan Biodiversitas dan distribusi Ikan Verh. Naturforsch. Ges. Basel 98, 51-57. di Danau Matano, Sulawesi Selatan (Prelimi- Kottelat, M., Whitten, T., Kartikasari, S.N., Wir- nary study: Biodiversity and distribution of joatmodjo, S. 1993. Freshwater fishes of Wes- fishes in Lake Matano, South Sulawesi). J. Ik- tern Indonesia and Sulawesi. Periplus Edi- tiol. Indonesia 2, 23-29. tions, Hong Kong. Haffner, G.D., Hehanussa, P.E., Hartoto, D. 2001. Munro, I.S.R. 1958. Handbook of Australian The biology and physical processes of large fishes, number 24. Australian Fisheries News- lakes of Indonesia: Lakes Matano and Towu- letter 17, 97-100. ti, pp. 182-192. In: The Great Lakes of the Munro, I.S.R. 1967. The fishes of New Guinea. world (GLOW) – Food web, health and integ- Department of Agriculture, Stock and Fish- rity (Munawar, M., Hecky, R.E. eds). Back- eries, Port Moresby. huis Publishers, Leiden. Roy, D., Docker, M.F., Hehanussa, P.E., Heath, Herder, F., Nolte, A.W., Pfaender, J., Schwarzer, D.D., Haffner, G.D. 2004. Genetic and mor- J., Hadiaty, R.K., Schliewen, U.K. 2006. phological data supporting the hypothesis of Adaptive radiation and hybridization in adaptive radiation in the endemic fish of Lake Wallace’s dreamponds: evidence from sailfin Matano. J. Evol. Biol. 17, 1268-1276. silversides in the Malili Lakes of Sulawesi. Saeed, B., Ivantsoff, W. 1991. Kalyptatherina, the Proc. Roy. Soc. Lond. B. 273, 2209-2217. first telmatherinid genus known outside of Hutchinson, G.E. 1957. A treatise on limnology. Sulawesi. Ichthyol. Explor. Freshwaters 2, John Wiley & Sons, New York. 227-238.

162 Saeed, B., Ivantsoff, W., Crowley, L.E.L.M. 1994. tropod Tylomelania (Gastropoda: Cerithioidea: Systematic relationships of atheriniform Pachychilidae) on Sulawesi, Indonesia and its fisheswithin division 1 of the series Atheri- biogeographical implications. Biol. J. Linn. nomorpha () with relevant his- Soc. 85, 513-542. torical perspectives. Vopr. Ikhtiol. 34, 1-32. von Rintelen, T., Wilson, A.B., Meyer A., Glaub- Sarasin, F. 1886. Durchquerung von Südost-Ce- recht, M. 2004. Escalation and trophic special- lebes. Verh. Ges. Erdkunde zu Berlin 23, 339- ization drive adaptive radiation of freshwater 356. gastropods in acient lakes on Sulawesi, Indo- Sparks, J.S., Smith, W.L. 2004. Phylogeny and bio- nesia. Proc. Roy. Soc. Lond. B 271, 2541-2549. geography of the Malagasy and Australasian Weber, M. 1913. Neue Beiträge zur Kenntnis der rainbowfishes (Teleostei: Malanotaenioidei): Süsswasserfische von Celebes. Ergebnisse Gondwanan vicariance and evolution in fresh- einer Reise von E. C. Abendanon in Celebes. water. Mol. Phyl. Evol. 33, 719-734. Bijdr. Dierk. 19, 197-213. von Rintelen, T., Glaubrecht, M. 2003. New dis- Weber, M., de Beaufort, L.F. 1922. The Fishes coveries in old lakes: three new species of of the Indo-Australian Archipelago. IV: He- Tylomelania Sarasin & Sarasin, 1897 (Gastro- teromi, Soleichthyes, Synentognathi, Perceso- poda: Cerithioidea: Pachychilidae) from the ces, Labyrinthi, Microcyprini. Brill, E.J., Lei- Malili Lake system on Sulawesi, Indonesia. J. den. Moll. Stud. 69, 3-17. Whitten, T., Henderson, G.S., Mustafa, M. 2002. von Rintelen, T., Glaubrecht, M. 2005. Anatomy The Ecology of Sulawesi. The Ecology of of an adaptive radiation: a unique reproduct- Indonesia Series IV. Periplus Editions, Singa- ive strategy in the endemic freshwater gas- pore.

Verh. Ges. Ichthyol. Bd. 5, 2006 163