ACTA BIOLOGICA SLOVENICA
LJUBLJANA 2019 Vol. 62, Št. 2: 45-66
ABS
What happened to the genus Gammarus (Crustacea: Amphipoda)
in Central Asia
Kaj se je zgodilo s postranicami Gammarus (Crustacea: Amphipoda)
v centralni Aziji
Boris Sketa*, Dmitry Sidorovb, Zhonge Houc
aDepartment of Biology, Biotechnical Faculty, University of Ljubljana, P.O. Box Ljubljana SI-1001, Slovenia bFederal scientific center of the East Asia terrestrial biodiversity, Far Eastern Branch of Russian Academy of Sciences, 100-let Vladivostoku Av. 159, Vladivostok 690022, Russian Federation cKey Laboratory of Zoological Systematics and Evolution, Institute of Zoology, Chinese Academy
of Sciences, Beijing 100101, China Correspondence: boris.sket@bf.uni-lj.si
Abstract: We examined the phylogenetic relationships in the unusually structured fauna of the amphipod genus Gammarus in Central Asia, trying to explain its causes. Mitochondrial DNA sequence data were used in MP and ML and Bayesian analyses to generate a phylogenetic hypothesis. All of the recently collected Gammarus spp. specimens belong to the G. lacustris aggregate, except scarce marginal populations of G. balcanicus and G. komareki aggregates. Although molecularly closely related, these. G. lacustris agg. populations are morphologically and ecologically equally diverse as are the numerous European Gammarus spp. Only few populations are morphologically close to the morphotype of G. lacustris. There is a more than 2,000 km wide gap between the eastern and western populations-species of G. balcanicus agg. Since Gammarus lacustris can be spread by birds, and its Asian relatives are molecularly significantly diverse, we suppose that it invaded Central Asia several times and from several sources. It most probably differentiated to diverse morphotypes after reaching Asia. The area of the cryophilic and less nomadic G. balcanicus could have been reached and conquered only by waterways. Its area was probably continuous in the past and divided by aridification during the Neogene. The strong historical aridifica-tions probably extirpated the probably higher Gammarus diversity in Central Asia, including G. balcanicus. This opened the area and its habitats for the reiterated invasions of the particularly euryoecious G. lacustris s.l., followed by its morphological and ecological diversification.
Keywords: Amphipoda, biogeography, diversification, Palearctic
Izvleček: Raziskali smo filogenetske odnose v nenavadno strukturirani favni po-stranic Gammarus v Centralni Aziji in jo skušali pojasniti. Mitohondrijske sekvence DNK smo uporabili v MP in ML analizah in tako oblikovali filogenetske hipoteze. Vse postranice rodu Gammarus, zbrane v zadnjih letih, razen nekaterih obrobnih in poredkih populacij agregatov G. balcanicus in G. komareki, pripadajo agregatu
46
Acta Biologica Slovenica, 62 (2), 2019
G. lacustris. Čeprav molekulsko tesno sorodne, so te populacije agregata G. lacustris ekološko in morfološko tako raznolike, kot so številne evropske vrste rodu Gammarus. Le maloštevilne populacije so morfološko blizu morfotipu G. lacustris. Med vzhodnim in zahodnim delom areala agregata G. balcanicus je tukaj 2000 km širok presledek. Ker lahko G. lacustris raznašajo ptiči in ker so njegovi azijski sorodniki molekulsko zelo raznoliki, predvidevamo, da so poselili Centralno Azijo večkrat in iz raznih smeri. Verjetno so se diverzificirali v različne morfotipe šele v Aziji. Bolj kriofilni in manj nomadski G. balcanicus s.l. je lahko osvojil svoj areal le vzdolž voda. V preteklosti je bil njegov areal gotovo zvezen in so ga razdelile šele neogenske aridifikacije. Ostre aridifikacije v preteklosti so verjetno iztrebile bogatejšo gamaridsko favno vključno z G. balcanicus. To je odprlo prostor in njegove habitate za ponavljane invazije posebej evriekega G. lacustris s.l., čemur je sledila morfološka in ekološka diverzifikacija.
Ključne besede: Amphipoda, biogeografija, diverzifikacija, Palearktika
Introduction
History of discovery and problems in taxonomy
Amphipods of Central Asia are insufficiently studied; distributional data are fragmentary and descriptions are mostly poor, often not mutually comparable. Only a few of the previously described Gammarus taxa have recently been confirmed with data that are more detailed. In addition, administrative and orthography changes in the near history added to the problems of the distributional data. Just as an example: the Martynov's (1935) 'Eastern Bukhara' is not part of the recent town Buhara, Uzbekistan, it is the eastern part of the extended Bukhara Emirate; its eastern part mainly included the central and western parts of the modern Tajikistan (Chibilev 1993).
Another problem concerns the numerous findings of 'Gammarus pulex' in Central Asia and adjacent regions. Its alleged first mention was by G.O. Sars (1901) from territories near Astana in Kazakhstan and eastern Mongolia, and then again by Sars (1903) from Lake Teleckoe, Altai. Much later, Martynov (1930b) reported four Amphipoda species for Teleckoe ozero, but G. pulex was not mentioned among them. However, material identified by Sars as G. pulex was described as new species, G. teletzkensis (see Martynov, 1930b). Surprisingly, G. pulex was listed by Martynov (1930b) from another Siberian lake and from the Biya river. Furthermore, Chevreux (1908), studying collections of prof. D.D. Pedashenko in 1906-
1908, reported G. pulex from a number of points located in the 'Lake Issyk-Koule, stream Gorge of Karakol, and Lake Chatyr-Koule' of Kirghizia. Afterwards, Martynov (1930a), based partly on the same collections of Pedashenko, described Gam-marus ocellatus Martynov 1930, and Gammarus bergi Martynov, 1930 from Lake Ysyk Kol (Issyk Kul). Once more later, G. ocellatus was indicated by Martynov and Behning (1948) from the high mountain lakes of the Hissar Range in Tajikistan. According to Martynov and Behning, 1948, G. ocellatus is a widely distributed species in the mountainous regions of Central Asia, occurring from Tajikistan to Tien Shan in the northeast and to the Himalayas in the southeast (Ueno 1934). The question of the existence of G. pulex in Asia gets significantly complicated with the impossibility to accurately identify a number of previous findings due to their superficial descriptions. In Turkmenistan, G. pulex was reported by Birstein (1945) from rivers Serlok and Firjuzinka (central Kopetdag); this is the only mention of G. pulex, where some morphological details show similarity with that European species.
The great majority of localities, treated by Martynov, are actually situated in southern Kazakhstan or very close to its border in Uzbekistan and Kyrgyzstan. The recently collected samples from Central Asia, which we could identify molecularly, are from Kyrgyzstan, E Uzbekistan, Tajikistan, and from Turkmenistan. They all virtually only contain members of the G. lacustris aggregate. They inhabit springs, streams, lakes, and caves, habitats,
Sket et al.: Genus Gammarus in Central Asia
47
Figure 1: Distribution map of Gammarus spp. in Central Asia and its surroundings. Asterisks with names - important cities.
Reversed triangles - molecularly not defined Gammarus spp.; rings - molecularly defined taxa of agg. lacustris; pale ribbons connect localities of G. cf. subaequalis-Garlyk, hatched ribbon connection with Martynov's G. sub-aequalis and G. turanus; numerals as in the text and Supplement table S2; squares - agg. balcanicus; diamonds (G. turcomanicus) - agg. komareki. Upright triangles - Sarothrogammarus group. Sarothrogammarus and G. balcanicus groups supposedly match with entire Asian parts of the group areas. Corresponding to Supplement S3.
Slika 1: Karta razširjenosti vrst rodu Gammarus v širšem območju Centralne Azije. Zvezdica z imenom - pomembnej ša mesta. Obrnjeni trikotniki - molekulsko nedefinirana Gammarus sp. (glej tabela v dodatku S2); obročki -molekulsko definirana taksa agregata lacustris; bledi pasovi povezujejo lokalitete G. cf. subaequalis - Garlyk; prekinj en pas, povezava z G. subaequalis in G. turanus Martynova; številke kot v tekstu in v tabeli v dodatku S1; kvadrati - agg. balcanicus; rombi (G. turcomanicus) - agg. komareki. Pokončni trikotniki - rodovna skupina Sarothrogammarus. Skupini Sarothrogammarus in G. balcanicus sta predvidoma cela azijska dela arealov. Se ujema s sliko v dodatku S3.
which in Europe would mainly be inhabited by G. pulex, G. fossarum, and G. balcanicus aggregates. Based on the available data and by analyzing the geographical distribution (see the map and species list, Supplement table S2), we made an attempt to conduct a comparative morphological analysis only for two new species described in another paper (Sidorov et al. 2018). Here, we are trying to establish, what happened with the rest of the Gammarus phylogenetic clades/aggregates, which had evidently inhabited the region in the past and only left behind some scarce relics.
Material and methods
Study area
Our study area was wider Central Asia, with exclusion of the Caspian in the West, and inclusion of Krasnojarsk in the East and North and of Islamabad in the south, but excluding the Tibetan plateau.
The areas sampled were as follows. In Turkmenistan, the surroundings of A^gabat (Ashgabat; where some localities were not accessible) and Köytendag area in the extreme NE. In Kyrgyzstan the lake Ysyk Köl (Issyk Kul) with surroundings
48
Acta Biologica Slovenica, 62 (2), 2019
and the SW of the country. In Tajikistan, the NW third of the country (sampled by C. Fiser) and the South-East (sampled by D. Palatov). In other countries, some scattered localities by different collectors. All localities and collectors are listed in the supplementary table (Supplement table S1). Geographical maps with locations of the sampling sites were constructed using an open source software Generic Mapping Tools, GMT 4.5.14. (Fig. 1).
Gammarids were sampled in springs and streams with a hand net. In lakes, small dredge and baited falls made of plastic bottles and left on deeper bottoms overnight were also used. Samples were fixed and stored in 96% ethanol with a small amount of glycerol added; such animals are appropriate for DNA analyses and still not too rigid for morphological studies. Single specimens were partly dissected and some muscle tissues or small body parts taken for DNA analysis. The rest of the bodies, including all taxonomically relevant parts, were retained as voucher specimens for subsequent morphological studies.
All samples were molecularly identified up to the level of phylogenetic clades, defined as aggregates. Note that our aggregates (abbreviations: aggs, agg) (Hou and Sket 2016) are not identical with ,groups' of Karaman and Pinkster (1977a, 1987), who explicitly define them just as morphologically (rather than phylogenetically) defined groups. Species identification was only taken as reliable if we could molecularly define morphologically appropriate samples from the topotype localities. If not, identification was marked as ,cf.'; such identification only shows a certain degree of similarity (i.e., morphotypes), allowed by the identification using available keys; it does not mean any degree of phylogenetic relatedness, let alone the taxonomical identity. In addition, a terrifying degree of homoplasy was noted in the morphological characters of Amphipoda (Hou and Sket 2011; Moskric 2016).
Molecular methods
Total genomic DNA was extracted from specimens using the Tiangen Genomic DNA kit. Mitochondrial cytochrome oxidase subunit I (COI) and nuclear 28S rRNA genes were amplified using
primers as in Hou et al. (2007). Chromatograms were proofed and edited using Sequencher 4.2 DEMO (Gene Codes Corporation, Inc). Sequences were aligned using Clustal X (Thompson et al. 1997) and translated with the invertebrate mitochondrial genetic code to check for stop codons to detect pseudogenes in MacClade 4.06 (Maddison and Maddison 2000).
The best-fit substitution model of HKY+I+r was selected for COI and GTR+I+r for 28S using jModelTest (Posada, 2008). The two fragments were analyzed separately to emphasize possible mitonuclear differences. The phylogeny was reconstructed under maximum parsimony (MP), maximum likelihood (ML), and Bayesian approach. MP analyses were performed using PAUP* 4.0b10 (Swofford 2002). All phylogenetically uninformative characters were excluded from the analysis, and gaps were treated as missing data. Heuristic searches were conducted using tree bisection reconnection branch swapping, with a limit of one million rearrangements for each replicate. Bootstrap support indices were generated based on 1,000 bootstrap replicates with ten random-addition sequences. ML analysis was conducted using RAxML 8.2.9 (Stamatakis 2014), starting with 1,000 rapid bootstrap replications followed by a thorough tree search. Bayesian analyses were conducted using MrBayes 3.2.6 (Ronquist et al. 2012), with 8,000,000 generations sampling every 100 generations. A majority consensus tree was constructed to estimate the Bayesian posterior probabilities, after 25% burn-in.
Molecular operational taxonomic units (OTUs or MOTUs) are defined as monophyletic clades with distinguishable morphological or geographic properties.
Results
Except for some representatives of the G. bal-canicus agg and the only member of G. komareki agg, all of the recently collected Gammarus samples from Central Asia are members of the G. la-custris agg (all sensu Hou and Sket 2016; Figure 1). However, for most of the previously described taxa, samples for molecular analyses have not been obtained. Unfortunately, the same also goes for
Sket et al.: Genus Gammarus in Central Asia
49
the endemic genus and species Issykogammarus hamatus Chevreux, 1908. None of our samples appeared to be morphologically identical to the lacustrine species G. ocellatus Martynov, 1930 or G. inberbus Karaman & Pinkster, 1977.
The phylogenetic analyses of COI and 28S markers produced very similar results. The COI dataset included a larger taxa assemblage, which allowed rough survey on the entire G. lacustris aggregate, while the 28S dataset was more useful for exploring the relationships between Central Asian species. The phylogenetic relationships within the G. lacustris agg. were largely unresolved. However, OTUs clustering for the more variable COI (Brown et al 1979) and for 28S was similar.
In its proximal part, the phylogram of the. G. lacustris agg. is split into a high number of long branches. Members of these long branches are also morphologically diverse, in part already described as species living throughout western Asia and the Mediterranean (Kyrgyzstan, Turkmenistan, Iran, Lebanon, Italy, Morocco).
Terminally, the tree is split into two clades, called here G. lacustrisA and G. lacustrisE. This is a morphologically homogeneous part of the tree, most of its MOTUs have been mainly recognized by different authors and eventually published as 'G. lacustris'. The clade G. lacustrisE is the European clade. The G. lacustrisA contains taxa from E Asia, from Himalayas (new data by ZH), some from Central Asia, but also from N America (USA, Canada), and few from Europe (Norway, Ukraine). The morphologically distinct G. bergi from Ysyk Kol make this clade morphologically more diverse, although the majority of members have been mentioned in literature as G. lacustris.
The widely dispersed G. cf. subaequalis-Garlyk and the lacustrine G. bergi are hardly distinguishable in the 28S tree, while they are broadly separated in the COI tree. In addition to that, G. bergi is morphologically distinct, while G. cf. subaequalis-Garlyk belongs to the G. lacustris morphotype (Figs. 2a, b).
50
Acta Biologica Slovenica, 62 (2), 2019
Sket et al.: Genus Gammarus in Central Asia
51
Figure 2: Phylogenetic tree of the aggregate Gammarus lacustris derived from maximum likelihood analysis of 28S (a) and COI (b) sequences. Bootstrap values (>70) are shown above branches. OTUs in both distal branches of the tree are strongly grouped and condensed. Marked subclades are from Central Asia. Slika 2: Filogenetsko drevo agregata Gammarus lacustris po analizi maximum likelihood markerjev 28S (a) in COI (b). Vrednosti bootstrap (če >70) so nad vejami. OTUji v obeh končnih vejah so močno združeni. Označene podveje so v Centralni Aziji.
52
Acta Biologica Slovenica, 62 (2), 2019
List of taxa and their distribution
The diversity of the Central Asian Gammarus can be mainly illustrated by the analysis of the phylogram of the G. lacustris aggregate (Fig. 2a and b).
In detail, taxa, attributed to Gammarus or Rivulogammarus with revised geographical data and with some comments where necessary, are listed in the supplementary data (Supplement table S2). To avoid the accumulation of weakly supported nomenclatural combinations, we are only directly citing Martynov's taxa, since any serious revision would demand previous molecular analyses. This means that we also preserved the original nomenclature with Rivulogammarus S. Karaman 1931 (one of the authors - (DS) - does not agree with such solution). Rivulogammarus is anyway an objective junior synonym of Gammarus (Stock, 1969). Note that virtually all of Martynov's taxa (and subtaxa) are based on single populations. Molecularly studied samples are marked here in text with '(DNA)'. Numbers in brackets after species names are designations of OTUs in the phylogram as well as in the supplement table S1.
Gammarus lacustris aggregate
The majority of Central Asian taxa and its genetic lineages are morphologically within the morphological diversity of the G. lacustris aggregate. Species of G. tianshan, Gammarus sp5, G. parvioculatus, G. troglomorphus, G. decorosus, Gammarus sp2, and Gammarus sp1 are distinctly branched from close to the base of the aggregate's phylogram. Being morphologically and molecularly so diverse, they might be the most ancient Central Asian Gammarus species with a long evolutionary history in the earlier biodiversity of the region. The species G. bergi and G. cf. subaequalis are nested in the clade of G. lacustrisA with short branches, which could be explained by recent diversification of the clade lacustrisA in Central Asia. Since G. bergi and G. cf. subaequalis-Garlyk inhabit special niches, namely a brackish lake and a streaming river, this may explain morphological differences from G. lacustris.
(1)	Close to the phylogram base, a loose clade of long-branched OTUs splits, including G. tianshan Zhao, Meng & Hou (432; syn. G. montanus Hou, Li & Platvoet, 2004), which is our only taxon from the lake Teleckoe (257) and other populations from Central Asia, and also a population from Ukraine (420) and the type population of G. varsoviensis from Poland. Martynov (1930b, 1935) described Gammarus teletzkensis Martynov, 1930, G. ocellatus subsp. angulatus Martynov, 1930, Gammarus korbuensis forma reducta Martynov, 1930 from the lake Teleckoe ozero (Altyn Kol), Altaj, Russia. We could not join our Gammarus sp. (DNA) OTU 257 G.teleck. to any of the nominal taxa described from that lake.
(2)	In the basal part of the tree, there is also a long-branched clade from Turkmenistan and Xinjiang. These are G. decorosus Meng, Hou & Li, 2003 (693; Xinjiang, China), G. troglomorphus Sidorov, Hou et Sket (513), and G. parvioculatus Sidorov, Hou et Sket (515), the latter ones from the extreme East of Turkmenistan. Gammarus troglomorphus Sidorov, Hou & Sket, 2018 (DNA) Turkmenistan E, Lebap Province, Garlyk, is one of the most extremely troglomorphic Gammarus species at all, eyeless and with long pereopods. It seems to be the only troglomorphic species and the only one with strongly carinate urosomites within the aggregate. It was found accompanied by a troglobiotic fish Triplophysa starostini (Parin, 1983) (syn. Noemacheilus s., Paracobitis s., Troglocobitis s.) in the collapse doline (karst sink) Suw Oyuk.
Gammarus parvioculatus Sidorov, Hou & Sket, 2018 (DNA) from Turkmenistan E, Lebap Province, Koytendag District, Khodzhapil (= Hojeypil) was found in two very different springs near Koyten. Although evidently an eutroglophile, it is a comparatively short-legged species with variable pigmentation and only slightly reduced eyes.
Thus, G. troglomorphus and G. parvioculatus are very morphologically different, not only at the expense of troglomorphy. On the other hand, they are molecularly sister species in both COI and 28S analyses. These molecular relations are weakly supported (with only 45%) in COI. In the 28S tree, the grouping of G. troglomorphus and G. parvioculatus is supported with a value of only 78. In the COI tree, their further sister
Sket et al.: Genus Gammarus in Central Asia
53
taxon is G. decorosus (OTU), an epigean animal from Xinjiang.
(3)	Phylogenetically more compact is a group of OTUs here named Gammarus sp. 2 (DNA), although it might be a group of independent species. They inhabit the lake Ysyk Kol (365, 368, at least up to 70 m deep) and springs (in Tokmok, Tyup, Karakol: 358, 362, 363, 376) between Karakol and Bishkek in the NE part of Kyrgyzstan. The group could not be identified either with Gammarus ocellatus ocellatus Martynov, 1930a, or with Gammarus inberbus Karaman et Pinkster, 1977b (described from the same lake at Koisara). All our gammarids from the SE part of the lake are provided with slight dorsal elevations on urosomites, while in G. inberbus, pleonites and urosomites are explicitly 'smooth, without dorsal elevations or excavations' (Karaman and Pinkster 1977). Moreover, our animals show less acute epimeron III and eye not as large as in Gammarus ocellatus ocellatus.
(4)	Phylogenetically comparatively isolated group with some branches around the base of the terminal bifurcation of the COI tree, taken from Tajikistan, Kyrgyzstan, and from the Indian Ladak. In the 28S tree, these OTUs are similarly grouped, just slightly differently positioned in the tree.
Gammarus sp. 1 (DNA), OTU 198 G Talas, from Kazakhstan, Dzhambul oblast', Taraz, r. Talas, is phylogenetically an isolated population, molecularly most related to species from Iran, Europe (Italy), N Africa.
Gammarus sp. 3 (DNA), OTUs 374, 375, is from Kyrgyzstan W, Khodzhaaryk, Abshirsai, Abshir Ata.
Gammarus sp. 4 (DNA), OTU 453 is from Tajikistan SW, Shaartuz district, oasis Chilychor.
Gammarus sp. 5 (DNA) OTU (454 TAD2) and G. tianshan are a comparatively isolated, basally split group, including also a specimen from Teleckoe ozero; the group is similarly positioned in both phylograms.
(5)	While the terminal clade G. lacustrisE is entirely European, G. lacustrisA includes few branches from N and E Europe (Norway, Ukraine), North America (USA, Canada), and eastern China, some isolated OTUs from E Asia and two important subclades from Central Asia. One of them is G. bergi Martynov, 1930 (DNA) (357, 361, 366, 367, 370), endemic to the brackish,
high elevation (1,607 m a.s.l.) lake Ysyk Kol. We found it up to 30 m deep on northern and southern sides of the lake (Bokonbaevo and Cholpon Ata). The species stands out of the G. lacustris agg. for its long and dense setosity of atennae I-II. It is roughly reminiscent to G. komareki, but four specimens analyzed molecularly clearly speak for its position in the G. lacustris agg. and away from the G. komareki agg. Moreover, in the tree it is surrounded by animals of the lacustris morphotype (i.e., morphologically identified as G. lacustris). An individual from deeper water (-30 m) has less setose (but still komareki-like) antennae. Terminal clades G. lacustrisE and G. lacustrisA loosely correspond to both allozyme races of Vainio and Vainola (2003).
The other important subclade here has been designated as G. cf. subaequalis-Garlyk (Sidorov et al. 2018), morphologically related to the following Martynov's taxa. Rivulogammarus subaequalis Martynov, 1935, with 4 subordinate taxa (and/or synonyms; Martynov's subordinate/subspecific taxa are: subspecies, natio, morpha, forma) is from S Kazakhstan and Tajikistan; Rivulogammarus turanus Martynov, 1935 with 6 subordinate taxa, which was found in Uzbekistan and S Kazakhstan. As Gammarus (Rivulogammarus) syriacus subaequalis, the former was supposed to be (Birstejn, 1948) 'very common, widely spread in Tajikistan' (springs at fortress Hisor (= Gissar, Hissar, r. Hanakinka, kishlak near st. Hanaka, kislak Hodzambio).
Thus, Gammarus cf. subaequalis-Garlyk (DNA) seems to be (a part of) the most important taxon of the Central Asia. The Martynov's G. subaequalis and G. turanus are morphologically mutually very close and may be phylogenetically identical with the molecularly studied OTU 516, described nowadays (Sidorov et al. 2018) as G. cf. subaequalis-Garlyk. The population from Garlyk (= Karlyuk), Koyten area in NE Turkmenistan resembles G. turanus particularly in the shape of pereopod V base, but it is very different in telson shape and setation. In telson and in uropod III dimensions and setation, it is closer to G. subae-qualis. Our taxon is very different from both in urosomal spines. Molecularly very close to 516 (Karabulak, Garlyk) are OTUs TAD3 (Zeravshan Valley), 261 (lake Alaudin), 369 (Ysyk Kol, -10 m), 423 (lake Iskandar Kol), 465 (lake Karakul,
54
Acta Biologica Slovenica, 62 (2), 2019
Pamir). Their phylogenetic relation (affinity) is also highly supported (90 BP). Considering the morphology and distribution, we suppose that these populations are really conspecific with one or both of the mentioned Martynov's species. Seven molecularly closely related populations are distributed within the area between the easternmost Turkmenistan, Tajikistan, and up to the eastern Kyrgyzstan with ca. 1,200 km distance between localities 516 and 369 (i.e., between Garlyk and Ysyk-Kol), while the Martynov's G. subaequalis and G. turanus are ca. 250 km away from this area to the NW. The phylogenetic relation between this taxon and G. troglomorphus, its very close geographical neighbor, is very weak.
If our supposition is correct, Gammarus subaequalis Martynov, 1935 is one of the most widespread species of the region, as indirectly already suggested by Birstejn (1948). Morphologically it is a lacustris-like species with a series of features occurring in phylogenetically diverse species within the genus Gammarus. The greatest morphological similarity outside Central Asia is with G. crinicaudatus Stock et al., 1998 from the southern Zagros region of Iran.
Gammarus balcanicus aggregate
Members of the G. balcanicus aggregate are morphologically comparatively easily recognizable. The reliability of such identification is, however, low, but in our case it does not contradict the few molecularly supported data. Characteristic of this aggregate is the poor setation of pereopods III-IV and of uropod III (Karaman and Pinkster 1987). In the latter, the spines along the exopodite outer margin are associated with only few and short finer setae.
Molecularly proven members in Central Asia are Gammarus takesensis Hou, Li & Platvoet, 2004 (DNA) from China, Xinjiang, Takes, and Gam-marus tastiensis (Hou, 2002, thesis not printed) (DNA) from China, Xinjiang, Yumin. The only morphologically recognized members of this aggregate are Gammarus brevipodus Hou, Li & Platvoet, 2004 from China, Xinjiang, Xinyuan, Gammarus angustatus Martynov, 1930a with two subordinate taxa from Russia, river Ob drainage (incl. Novosibirsk), Gammarus pellucidus Gur-
janova, 1930 from Russia, Krasnojarsk, Gammarus (Rivulogammarus) balcanicus Birštejn, 1950 from Russia, river Listvjanka, Jurginskij rajon, Kem-erovska oblast' and Rivulogammarus spinulatus Martynov, 1935 from Kazakhstan, E of Almaty.
Thus, we can confirm the presence of the eastern assemblage of the agg. balcanicus along the E Kazakhstan-Xinjiang (China) border and in the adjacent areas of Russia. No member of agg. balcanicus could be detected in Iran. The closest western agg. balcanicus members are probably (not DNA-confirmed) in Asia Minor - G. orientalis (s. Karaman 1934) (Karaman 2017; in Central Turkey), G. burduri S. & G. Karaman, 1959 (SW Turkey), and in Azerbaijan - G. alarodius Deržavin, 1938 (in Nakhichivan, river Araks), G. talyschensis Deržavin, 1939 (S Azerbaijan).
Gammarus komareki aggregate
The Gammarus komareki aggregate is represented by Gammarus turcomanicus Birštejn, 1945 (DNA) (syn. G. (Rivulogammarus) balcanicus turcomanicus Birštejn 1945) at the edge of the area, in Turkmenistan, rivers A^gabatka, Ki^inka, Kara-su, Zolotoj ključ, Qulinka, Sakiz-jab around A^gabat (Birštejn 1945). Molecular analysis (new data) shows the appurtenance of the spring population of Qulinka to agg. komareki. Although described as G. balcanicus ssp., its morphology is not in disagreement with such solution. This is the easternmost known population of the G. komareki agg. Further members of this aggregate are present in Iran (Stock et al. 1998, Zamanpoore et al. 2011), in Caucasus (Birstein 1933, Behning 1940, Behning and Popowa 1947), in E Europe, and probably in Asia Minor (Alanyali et al. 2011).
Discussion
Ruling of the G. lacustris aggregate in Central Asia
Our gammarid samplings clearly show that - beside the Sarothrogammarus group in high mountains - we could only find members of G. lacustris in Central Asia. However, the core area is sparsely surrounded by members of the
Sket et al.: Genus Gammarus in Central Asia
55
G. balcanicus aggregate and by G.komareki in the West. We were also not able to phylogenetically classify the Martynov's (1935) taxa, originating mainly in southern Kazakhstan. Members of this aggregate even inhabit aberrant habitats, such as a brackish mountain lake and subterranean waters, the latter highly morphologically modified. The G. lacustris relatives here are morphologically very diverse, not as unified as the populations in Central and Northern Europe, which could be identified as a unique G. lacustris ,morphospecies' (Karaman and Pinkster 1977, Alther et al 2016).
Some additional taxa had been described, for which we find no close phylogenetic relations to our molecularly studied samples. These are G. ocellatus Martynov, 1930a with three subordinate taxa from Ysyk Kol and Teleckoe ozero, indicated later by Martynov and Behning, 1948 for a wider high mountain area in Central Asia. Farther R. truncatus Martynov, 1935 with 2 subordinate taxa from Uzbekistan and Tajikistan; Gammarus gracilis Martynov, 1935 from Kazakhstan, eastern Karatau; Gammarus angusticoxalis Martynov, 1935 from Kyrgyzstan; Rivulogammarus brevicornis Martynov, 1935 from Kazakhstan S, Karatau, springs at rivulets Kichkin, Ishfan-Mazar, Martynov (1935): 'A separate species, but close to R. turanus\
Sidorov described Gammarus montaniformis Sidorov, 2012 from Kazakhstan, eastern Tien-Shan, and Gammarus alius Sidorov, 2012 from Kyrgyzstan.
Rivulogammarus hirsutus Martynov, 1935 (syn.: Rivulogammarus brachyurus Birstein, 1935, Gammarus birsteini Karaman et Pinkster, 1977) with 3 subordinate taxa from S Kazakhstan is morphologically very different and taxonomi-cally intriguing. Antennae resemble G. komareki or G. bergi; according to Martynov (1935), it is close to G. turanus; the uropod III endopodite is shortened to 50% of exopodite or less. The uniquely long dorsal setae on urosomites, in Martynov's morph 'hirsutissimus' even on pleonites, make its belonging to agg. lacustris or agg. komareki unlikely, but not impossible.
Birstejn (1945) considered G. hirsutus Martynov, 1935 a junior synonym of G. brachyurus Birstejn, 1935; since the name G. brachyurus appeared to have been preoccupied (Karaman and Pinkster 1977), the name hirsutus may nev-
ertheless be used. The Karaman and Pinkster's (1977) G. birsteini has been illustrated according to specimens from Turkey; since it lacks those characteristic urosomal setae, it is evidently a species per se.
Issykogammarus hamatus
A specific problem is Issykogammarus hama-tus Chevreux, 1908. It is an aberrant taxon of an uncertain position. It was obtained from different depths of the lake Ysyk Kol, Kyrgyzstan NE. It was described as a separate genus (Chevreux 1908), later attributed to principally Bajkalian acanthogammarids as a 'Baikalian Escapee' (Barnard and Barnard 1983). But, if we only overlook the recast of the pereopod coxae, the species fits easily into the diversity of the aggregate lacustris or even very close to the morphotype lacustris. The evolutionary mutability of the mentioned coxal armament in amphipods can be best documented by the marine genus Epimeria (Rauschert and Arntz 2015) or by the general body shape variability in Bajkalian gammarids, nested phylogenetically within Gammarus (Hou et al. 2016). Unfortunately, a molecular analysis was impossible and the old-fashioned presentation of the morphology, conducted even mainly on females, does not allow for a serious reevaluation of the species' taxonomy and phylogeny.
Gammarus pulex and its aggregate
Gammarus pulex and its aggregate in Central Asia remain a mystery. The only seemingly reliable data on G. pulex are those by Birstejn, 1945 from the rivers Sherlok and Firjuzinka near A^gabat (Ashgabat) (central Kopetdag, neighbor to G. turcomanicus). Explicitly mentioned are curled posterior setae on pereopod III (not so in pereopod IV) and widened and densely setose flagellar articles of antenna II; we could only hardly expect these striking characters of G. pulex pulex (Linnaeus) in G. lacustris relatives. Nevertheless, both species are very similar and the Asian members of the aggregate lacustris are explicitly diverse.
56
Acta Biologica Slovenica, 62 (2), 2019
Martynov (1930a, 1935) is persistently mentioning 'G. pulex' for lentic(!) waters in Central Asia without giving any further data. One could suppose that he was misled by the similar G. lacustris, which he surprisingly did not mention at all. According to Karaman (1984), 'almost not any of these (cited here) populations belong to the G. pulex' - two pages of citations are given, more than 60 populations for the entire Asia between the Caspian and the Pacific shores, with exception of its Indian (Southeast Asia, Indonesia) region.
Ecology and ethology
Gammarus lacustris has been characterized as an euryhaline, eurythermal, and euryoxibiotic species-ubiquist (Bazikalova 1941, Birstein and Taliev 1946, Bekman 1954, Pirozhnikov 1932, 0kland and 0kland 1985). In the vast territory of its rather well documented geographic distribution (see also Straskraba 1966), the morphotype 'Gammarus lacustris' inhabits various water bodies: limnic and running waters, cold and warm, fresh and salty, rich in oxygen and hypoxic. In the Trans-Urals and in the Middle Urals, it inhabits soft fresh water with very low salt content (dry residue 60-77 mg/l) and brackish water bodies, for example in the Chelyabinsk region of the lakes Chebarkul and Cancul with dry residue 8,83 g/l (Dexbach 1952). In the Baikal region, it inhabits low-mineralized water bodies with a content of dry residue from 25 mg/l to 152 mg/l (Bekman 1954). In the lakes of the West Siberian lowland, it lives in water bodies whose salinity reaches up to 10% (Balabanova 1941, 1949).
A number of authors (Buschkiel 1931; Haem-pel 1908) pay attention to the concentration of calcium ions in the water. Therefore, Wundsch (1922) managed to show the mosaic distribution of Gammarus in the rivers of the right bank of the Rhine, where he stated the absence of Gammarus if the content of calcium ions was less than 9 mg/l. Experimentally, Bekman (1954) found out that the normal ontogenetic development of the Baikal Gammarus was disturbed in waters containing less than 7 mg/l of calcium.
In Central Asia, members of the. G. lacustris agg. inhabit all types of perennial aquatic habitats.
They have mostly been found in springs, while rivers were sampled occasionally. Since temperatures of ground waters (and springs) normally approach mean yearly temperatures of the area, we can expect the temperatures of springs to range from ca. 18 °C (at A^gabat, at 300 m a.s.l.) in the foothills to 7.5 °C in higher areas (around Ysyk Kol at 1,650 m a.s.l.). The terrain is mainly magmatic, dictating low calcium content in water; there are smaller karst i.e., carbonate areas (particularly in SE Kyrgyzstan and NE Turkmenistan), where we may expect hard waters. An even higher level of mineralization is reached in the lake Ysyk Kol itself, where water is brackish, with 6 g/l chloride/ sulfate/sodium/magnesium-based salinity (Aladin and Plotnikov 1993, Baetov 2005).
Martynov (1935) mentioned stagnant waters as a common dwelling place for G. pulex, but we only found some normal G. lacustris agg. members in some smaller lakes. Unfortunately, Martynov explicitly states only ponds close to Bishkek, the area where we found G. cf. lacustris. Several gammarids have been described from the lake Teleckoe, as well as from Ysyk Kol. Some species may be present in the depths of some tens of meters. Issykogammarus hamatus and Gammarus bergi are definitely endemic, while we cannot confirm this for G. inberbus or G. ocellatus. I. hamatus exhibits characteristics of the 'lacustrine morphology' (Hou et al. 2016), while other species from this lake are morphologically trivial.
G. troglomorphus is an evidently highly specialized troglobiont. G. parvioculatus is present in both, springs and underground; its only troglomorphies are slightly smaller eyes and integumental depigmentation in some individuals. Some other troglobiotic amphipods are also Tadzocrangonyx spp. (Sarothrogammarus group of Gammaridae) and Bogidiella spp. (Bogidiellidae; found in Turkmenistan).
Gammarids are not provided with ontogenetic stages that were particularly suited for lateral dispersal. However, appropriate behavior may enable the adult to disperse laterally, outside the water. Segerstrale (1954) has shown that the ethology of G. lacustris enables it to take the advantage of transportation by birds (ornithochory). The author could experimentally establish that this is not a general ability of gammarids, since G. pulex was behaving differently. Therefore, we also
Sket et al.: Genus Gammarus in Central Asia
57
do not know whether this is only an ethological adaptation or is it also morphologically dependent (supported). This capability may also be either inherent in the whole aggregate or limited to the lacustris morphotype. Only the remotely related Crangonyxpseudogracilis Bousfield, 1958 was later a subject for a similar experiment (Racha-lewski et al. 2013). In South Dakota, Hyalella azteca (Saussure) was found in duck's plumage (Rosine 1956). In Canada, the same amphipod was found in the fur of muskrat and beaver, while numerous G. lacustris were found in the fur of beaver (Peck 1975). We have no information about the respective behavior of other gammarids. We do not know which morphological characters contribute to the ability to hang on birds and we do not know the ethological differences and similarities between all of the numerous members of the G. lacustris aggregate. But, the exceedingly large distribution area of the G. lacustris agg. (and even of the morphotype 'G. lacustris') shows that this group is extraordinarily capable of dispersal. It either dispersed in morphologically diverse types, or it dispersed as the morphotype lacustris, while its populations morphologically diversified only after reaching their recent dwelling areas.
The supposed historical biogeography of Gammarus in Central Asia
Gammarus in Eurasia (Palearctic). The Eurasian continent is inhabited by phylogenetically and geographically very distinct gammarid groups (Hou et al. 2011, Hou and Sket 2016). The Asian continental freshwaters east of Ural and north of 55 oN are virtually devoid of Gammarus and mostly inhabited by Pallasea and Synurella, which regularly significantly dominates in tundra and forest-steppe in the northern biotopes. The few known populations (Derzavin 1923, Kurenkov and Mednikov 1959, Labay 1998) are either not attributable to an aggregate, or belonging to G. lacustris agg., with the exception of G. koreanus Ueno, 1940 in Sahalin, probably belonging to the Oriental aggregate. It is also necessary to take into account the fact that G. lacustris was purposely introduced almost everywhere in the lakes in the Central Russian Upland, Siberia and even in Kamcatka during the Soviet times. These doings
have not always been properly documented. The eastern portion of China and areas NE from it (Japan and Korea, Russian Far East) are inhabited by the Oriental aggregate of Gammarus; Lake Bajkal has its own endemic fauna, nominally of diverse genera (and even families, e.g. Tahteev 2000), which are in fact phylogenetically members of Gammarus and probably sister to the aggregates balcanicus and lacustris (Hou and Sket 2016). In the West, the European aggregates G. pulex, G. fossarum, and G. roeselii do not cross the eastern borders of Europe. Two SE European aggregates are extending towards Central Asia. The. G. komarekiagg. is richly represented in Iran and with G. turcomanicus Birstejn, 1945 (syn. G. balcanicus turcomanicus) in S Turkmenistan at A^gabat (Ashgabat). The repeatedly mentioned G. pulex (e.g., Martynov 1935) remains a total mystery regarding Asia; molecularly, the morphologically similar aggregates G. pulex, G. fossarum, and G. roeselii could have only been verified for the European territory, with inclusion of Georgia (Hou et al. 2014, Supplement table S1).
The G. balcanicus aggregate occupies the E and SE Europe with some probable, but not molecularly proven representatives between Europe and the Caspian (see above). Molecularly proven members are only present in Ukraine and further west (Hou et al. 2014). In SE Europe, G. balcanicus agg. populations, cited mainly simply as 'G. balcanicus", are evenly distributed, although populations of three other aggregates are also present there. Karaman and Pinkster (1987) mention more than 1,000 samples (of the supposed unified species!) in their area, Poland to SW USSR.
However, an additional detached area exists with a small number of reliable agg. balcanicus members (Hou and Sket 2016) along the border between Xinjang and Kazakhstan, and (not molecularly proven) in the adjacent areas of southern Sibir' (Siberia, Russia). This results in more than 2,000 km wide gap between both parts of the entire agg. balcanicus distribution. This gap includes the heart of Central Asia. In recent years, we sampled ca. 40 springs and clear lakes in Central Asia, typical habitats of 'Gammarus balcanicus', but not one sample contained balcanicus-liike animals. It was molecularly proven that they all belong to the G. lacustris agg. Even if some undiscovered populations of the G. balcanicus agg. do exist
58
Acta Biologica Slovenica, 62 (2), 2019
there, they must be very scarce, of an evidently relictual character.
Thus, the G. balcanicus agg. is only approaching Central Asia marginally in the NE. Such situation in a sedentary and highly endemic group, as agg. balcanicus is, can only be explained by a temporary omnipresence of the aggregate in the past. Thus, we suppose that in the past, the aggregate members used to be continuously present between Europe and NW Xinjiang, and in the South of the West Siberian Plain.
In Europe, a comparable area would be richly populated by gammarids of aggs. G. pulex, G. fos-sarum, and G. balcanicus, while the morphotype 'G. lacustris' is the only representative of the G. lacustris agg. There, it is mainly restricted to lakes (Alther et al. 2016; Vainola et al. 2017). In Southern Europe and the Mediterranean, G. lacustris agg. is also present in springs and streams and may be represented by morphologically different species (e.g., G. italicus, G. elvirae). In Italy, only the G. lacustris aggregate is present along the peninsula and in the Sardegna (Sardinia) island. The others are limited (with a small exception) to the continental North of the country, while Sicilia (Sicily) and the French Corse (Corsica) are devoid of a freshwater Gammarus. Some populations of 'G. italicus' from the central Italian lakes (later separated as G. elvirae Iannilli & Ruffo, 2002) are even 'very similar to G. lacustris' (Kara-man 1993). However, all these areas are richly inhabited by the related Homoeogammarus (syn. Echinogammarus p.p.).
Within the peninsular part of Italy (Stoch and Ruffo 2003), agg. balcanicus is present in two closely positioned UTM quadrates, the agg. lacustris in ca. 45 quadrates of 10x10 km (G. elvirae ca. 40 quadrates, G. lacustris 5 quadrates). The situation is similar in the Alpine and Scandinavian areas of central and northern parts of Europe, where agg. pulex is somehow limited, while the others are close to absent, and G. lacustris agg. (by G. cf. lacustris) is mainly common in lakes (Fryer 1953; Segerstrale 1954, Alther et al. 2016). On the contrary, in Slovenia, which lies at the SE edge of the Alps, we registered G. fossarum agg. from 120 localities (but hundreds of noted remain unregistered), G. roeselii agg. from 13 in the East, G. balcanicus agg. from 6 scattered localities in
the West, and G. lacustris agg. only in one lake in pre-Alps (our own data).
Phylogeography of Gammarus taxa in Koytendag
Very instructive, predicting the next surprises in the Gammarus fauna of Central Asia, are the relations among the eastern Lebap (Koyten-Garlyk (=Karlyuk)), extreme East of Turkmenistan) species. Two cave-related species (G. troglomorphus, G. parvioculatus) are morphologically the most dissimilar species of Central Asia, and among the most divergent species pairs within Gammarus at all, which is only to some extent the result of troglomorphy. As exposed above, both cave-related species are extremely morphologically different, and molecularly loosely, but firmly related. Their localities are ca. 40 km apart, while their probably (with low support again) closest relative is G. dec-orosus Meng, Hou & Li, 2003 from ca. 2,000 km away in Urumqi, Xinjang, which is not inclined to inhabit subterranean waters. All of the mentioned species represent relatively long branches in the proximal part of the aggregate phylogram, with contrast to the basal G. tianshan and others.
The third species from the Koyten-Garlyk area is clearly epigean, geographically the closest neighbor of G. troglomorphus at Garlyk, but evidently broadly dispersed. We tentatively named it G. cf. subaequalis-Garlyk (Sidorov et al. 2018). See above for some more details. It is a member of a small clade of 7 samples (our OTUs 261, 423, 465, 516, 369, 359, S135) in the tree from the 1,000 km long stripe between Garlyk (= Karlyuk) in Turkmenistan, Tajikistan, and Ysyk Kol area in Kyrgyzstan. Only future field sampling can enable us to tell whether these series of populations is conspecific with the named Martynov's taxa. The samples of G. cf. subaequalis-Garlyk are molecu-larly close or nearly identical, and as a group very close to the branchlets of the terminal part of the tree, its clade lacustrisA, including numerous samples from Eurasia and North America. The endemic G. bergi Martynov from the lake Ysyk-Kol is also included. Only ca. 300 km to the NW from the central part of this area are Martynov's (1935) localities of G. subaequalis and G. turanus, which are morphologically similar and might be
Sket et al.: Genus Gammarus in Central Asia
59
both, mutually conspecific and conspecific with the mentioned group.
The climatic past possibly influencing the distribution of gammarids in Central Asia
Central Asia has been the area of the highest aridity because of the rain shadow caused by Tibetan Plateau and some other mountain ranges (Caves et al. 2014). This aridity was increasing during the Cenozoic. The precipitation should have 'more than halved' over the Neogene. It is likely that this aridity extirpated the majority of the previously evidently existing agg. balcanicus populations; this opened the territory for the invasion of the vagile and euryoecious G. lacustris agg., which may have happened several times during short climatically less adverse periods.
The Central Asian taxa are very diversely positioned in the phylogram; some have branched off close to the tree root, while some are members or sisters of the most distal branches. One has to bear in mind that the distribution area of the entire G. lacustris agg. covers the entire area of the genus Gammarus and even crosses the other gammarid species (Hou et al. 2011, Hou and Sket 2016, Vainola et al. 2017). To a great degree, it occurs even in the shape, morphotype of G. lacustris, and is therefore often morphologically identified and published as ' Gammarus lacustris'.
Such distribution can probably be explained by the ability of G. lacustris to be transported by birds, as shown by Segerstrale 1954, hand in hand with the above-mentioned ecological universality of 'G. lacustris'. Waterfowl from the North (here Jamal and Tajmyr,) migrate in a wide front, but birds often fly along the coast of the seas and along large rivers (here Irtys) and Central Asian lakes (Thompson et al. 2001, Snitnikov 1980, Kydyraliev 1990). Therefore, the populations of amphipods can be steadily replenished in their basins. The erratic spreading of this group is also supported by the fact that molecularly very close populations occur in very distant localities, including the sisterhood between the populations from North America and Ysyk Kol.
The morphological diversity of the G. lacustris agg. members shows its evolutionary plasticity and
a possibility that its populations (and future species) could have distributed at different times, but every one probably still in the 'original' lacustris morphotype, i.e. conquistador's appearance. Thus, numerous subsequent colonizing groups formed lacustris-morphotype populations and diversified only after colonization of new territories, in this case in Central Asia. Such diversification might be a consequence of the enforcement between at least some lake (e.g., in Ysyk Kol) populations. No case of species sympatry within the agg. lacustris has been recognised outside Central Asia.
Precipitation quantity had been fluctuating in the near past (Wolff et al. 2017), and it was probably so during the whole Cenozoic period as well. Therefore, the G. balcanicus agg. could have spread its area from the West into Central Asia and beyond either before the onset of aridi-fication, or also during a longer or shorter wetter period within its course.
The only additional gammarids (i.e., non-Gammarus) in Central Asia are members of the eastern (Asian) clade of the Sarothrogammarus group of genera (Hou and Sket 2016, Barnard and Barnard 1983), including Tadzocrangonyx spp. They are ecologically specific, living mainly in springs and to some extent in the subterranean, between 2,000 and 3,000 m a.s.l., and hardly ever below 1,000 m a.s.l. They are limited to the SE part of Central Asia: Uzbekistan - Tajikistan - N Afghanistan - N Pakistan. It is possible that different ecological regime in high elevations prevented extirpation of the Sarothrogammarus gammarids.
Conclusions
(1)	Recent samplings of gammarids in the core area of Central Asia, approximately between E Turkmenistan - N to E Kirgyzstan - Tajikistan, which we were able to study molecularly, delivered only members of the G. lacustris aggregate. The area is sparsely surrounded by members of the G. balcanicus aggregate. Members of the related Sarothrogammarus group of genera are only present in springs and mainly in higher elevations.
(2)	The G. lacustris aggregate appeared to consist of species with high phenotypic and evolutionary plasticity, resulting in a morphologically diverse and geographically widely distributed
60
Acta Biologica Slovenica, 62 (2), 2019
taxon. Its distribution area is broader than that of the rest of the genus Gammarus. In addition, within the genus' area, it is present where other species are absent - including in Central Asia.
(3)	While the G. balcanicus relatives and other Palearctic groups are evidently only able to disperse along streams, the G. lacustris relatives are able to disperse laterally outside water, by ornitochory. We do not know how general this ability is within the aggregate.
The Gammarus assortment in C Asia consists of species from all parts of the phylogram. They evidently immigrated in part in different times from far abroad, but some speciation must have happened within the area, to a smaller extent probably even within the lake Ysyk Kol (Issyk Kul) alone.
No members of the generally European clade G. lacustrisE were found in Central Asia, but Central Asian species are in close relationship with other species from Europe, Asia, and North Africa - and even N America.
(4)	The absence of other Gammarus groups in this wide arid area and its paleoclimatic history indicates that the - probably originally richer -amphipod fauna in the lower regions could have been extirpated in particularly arid periods. The G. lacustris relatives could survive to some extent for their euryoecity and be restored by a steady import and immigration. The G. balcanicus populations could only survive sheltered in springs in marginal positions of the area.
So, a great gap in the Gammaridae diversity appeared in Central Asia
(5)	The morphological diversity of the G. lacustris agg. in Central Asia and also in general nearly equals the diversity of the European Gammarus spp.
The generally very poor subterranean fauna in Central Asia is represented by the troglobiotic and highly troglomorphic G. troglomorphus, and by the very stocky and still occulated eu-troglophile G. parvioculatus.
Povzetek
Ob raziskovanju podzemeljske biotske pestrosti v Koytendag, Turkmenistan, v okviru misije britanskega Royal Society for the Protection of
Birds, smo opazili nenavadno sestavo favne postranic rodu Gammarus. To nas je napeljalo na podrobnejšo raziskavo postranic v Centralni Aziji.
Postranice (Amphipoda) v Centralni Aziji so pomankljivo raziskane. Opisi so večinoma skromni. Velike težave so z geografskimi podatki. Zaradi administrativnih in pravopisnih sprememb so stara krajevna imena spremenjena, nova pa neustaljena. Tako npr. staro ime 'Eastern Bukhara' ni del današnjega mesta Buhara, Uzbekistan, temveč vzhodni del obsežnega Emirata Buhara, ki je vključeval predvsem osrednje in zahodne dele sodobnega Tadžikistana. Druga skupina problemov zadeva številne najdbe domnevnega Gammarus pulex v Centralni Aziji, ki tam, kot kaže, sploh ni prisoten. Po drugi strani pa Martynov (1935) sploh ne omenja G. lacustris, ki je s sorodniki prevladujoča skupina vrst. Martynov je okoli leta 1935 napisal nekaj člankov in obsežno raziskavo z dokaj površnimi in skromno ilustriranimi opisi, večinoma iz (danes) južnega Kazahstana. Izgleda, da je njegov Gammarus subaequalis Martynov, 1935 zelo podoben ali identičen nekaterim našim vzorcem in zastopa tam najbolj razširjeno vrsto.
Zdaj zbrani vzorci so večinoma iz Kirgizije, vzhodnega Uzbekistana, Tadžikistana in iz Turkmenistana. Ti omogočajo tudi molekulsko analizo, težko pa jih je povezati s taksoni, ki so jih postavili Martynov, Birštejn in Schellenberg. V dodani tabeli je pregled vseh že prej objavljenih podatkov.
Naša raziskava zajema v Turkmenistanu A^gabat z okolico in Köytendag na skrajnem sverovzhodu; v Kirgiziji jezero Ysyk Köl (Issyk Kul) z okolico ter jugozahod dežele; v Tadžikistanu kraje na severozahodu in jugovzhodu. Vse lokalitete in zbiralci so našteti v tabeli. V največji meri smo vzorčili izvire in jezero, le malo pa reke. Živali smo molekulsko identificirali do agregatov, 'skupin vrst', ki se sploh ne ujemajo z izključno morfološko določenimi skupinami Karamana in Pinksterja. Za določitev vrst se zanesemo le na vzorce iz topotipskih lokalitet, v kolikor so tudi morfološko primerni. Pri vseh drugih veljajo imena le pogojno in so označena s 'cf.'.
Za molekulsko analizo smo vzeli majhne dele živali, preostanek takšnih osebkov smo spravili za poznejšo natančnejšo morfološko in taksonomsko obdelavo. DNK smo izolirali s pomočjo Qiagen kita in pomnožili mitohondrijski COI ter jedrni 28S gen s primeri, ki so opisani v Hou et al. (2007). S
Sket et al.: Genus Gammarus in Central Asia
61
programom jModelTest (Posada, 2008) sta bila izbrana modela susbstitucij HKY+I + G za COI in GTR + I + G za 28S. Filogenijo smo rekonstruirali z metodama največje varčnosti (MP) in največjega verjetja (ML). Podporo razvejitev smo preverjali z bootstrap analizo (metodo vezanja).
Razen maloštevilnih pripadnikov agregata balcanicus in vzorca iz agregata komareki smo vse vzorce molekulsko razkrili kot pripadnike agregata lacustris. Filogenetske analize s COI in 28S so dale zelo podobne rezultate. Žal nismo imeli na voljo aberantnega endema Issykogam-marus hamatus Chevreux.
Od proksimalnega dela filograma za agregat lacustris se cepijo številne dolge veje. Tudi tak-soni s teh vej so morfološko raznoliki, deloma opisani kot samostojne vrste, ki živijo v zahodni Aziji in v evropskem Mediteranu. Na vrhu je drevo razcepljeno v dve veji. Taksoni na njih so si podobni, v literaturi so največkrat označeni kot G. lacustris; zelo drugačen je jezerski G. bergi. Ena teh vej je evropska; druga iz severovzhodne Evrope sega po eni strani v Severno Ameriko in po drugi v obsežne dele Azije. Gammarus cf. subaequalis-Garlyk je zelo razširjen po Centralni Aziji in je po morfotipu lacustris.
Pestrost centralnoazijskih gamaridov torej lahko ponazorimo z analizo filograma za agregat lacustris. Velja upoštevati, da imajo skoraj vsa taksa in subtaksa Martynova za osnovo le posamezne populacije. V drevesu agregata lacustris je še mešanica vrst, ki so, in takih, ki niso prisotne v Centralni Aziji. Izgleda, da je najbolj razširjen G. cf. subaequalis-Garlyk. Med dolgimi vejami pri bazi drevesa omenimo nedavno odkriti podzemeljski vrsti iz Turkmenije. G. parvioculatus je očitno evtroglofilna žival z malce zmanjšanimi očmi, sicer pa bolj kratkonoga v primerjavi z večino povsem površinskih vrst. Nasprotno pa je G. troglomorphus menda najbolj troglomorfna vrsta svojega rodu. Ti vrsti iz severovzhodne Turkmenije sta si filogenetsko sestrski, obenem pa med seboj morfološko najbolj različni vrsti v Aziji; različni si nista le na račun troglomorfnosti. Njima najbližji sorodnik je povsem površinska vrsta G. decorosus iz kitajskega Xinjianga. Od že znanih vrst v jezeru Ysyk Kol smo nekatere našli in molekulsko določili, nekaterih pa nismo našli. Endemna jezerska vrsta G. bergi nekoliko spominja na G. komareki, a jo mole-
kulski podatki zanesljivo postavljajo v agregat lacustris.
Issykogammarus hamatus je endemna jezerska vrsta. Jezero je visokogorsko (1,650 m n.m.) in somorno. Žival ima bočne trnaste izrastke, kar je značilno le za vrste iz starih jezer. Zato so ga uvrstili med živali tega tipa. Vendar pa se po vseh drugih morfoloških značilnostih dokaj ujema z G. lacustris.
G. turcomanicus je bil kot G. balcanicus turcomanicus napačno uvrščen. Pri A^gabatu je najbolj vzhodno nahajališče (te vrste in) agregata komareki. Populacije agregata G. balcanicus se zadirajo v naše območje iz južne Rusije (Sibirije) proti jugozahodu, v osrednjem delu Centralne Azije jih ni, pojavijo pa se spet na jugovzhodu Evrope. Med obema deloma areala je pas brez njih, širši kot 2,000 km. To so živali, vezane na stalne, hladnejše vode. Agregat pulex za Azijo je in ostaja misterij. Starejši avtorji ga omenjajo pogosto, večinoma brez opisov, zdaj pa ga nismo našli in te vrste v Aziji verjetno ni. Od drugih rodov v Srednji Aziji živi rodovna skupina Sarothrogammarus; tukajšnja podskupina je vezana na višje lege in jo široka vrzel (ca. 2,500 km) ločuje od mediteransko-atlantske podskupine rodov in vrst.
V Centralni Aziji smo našli gamaride v vseh tipih voda. Tako v mehkih, kot v trdih in v somornici, v potokih, bistrih mlakah, jezerih in izvirih, od ca. 300 do nad 2,000 m n.m. V geološki preteklosti je Srednja Azija utrpela aridifikacije, ki so nedvomno povzročile siromašenje voda ter izumiranje manj evriekih gamaridov. Zlasti v vla-žnejših obdobjih je bilo tako olajšano poseljevanje zelo evriekemu G. lacustris.
Za razliko od srednje in severne Evrope, kjer je agregat lacustris zelo vezan na jezera, se tukaj pojavlja v vseh tipih voda. Gamaridi nimajo stadijev, posebej prirejenih za razširjanje, pač pa so za G. lacustris ugotovili zmožnost razseljevanja s ptiči. Domnevamo, da so se na ta način razširjali v Srednjo Azijo in po njej. Na ta račun domnevna vrsta G. lacustris (torej morfotip tega imena) zaseda območje, ki presega siceršnji areal rodu Gammarus in celo družine Gammaridae.
Rod Gammarus v Srednji Aziji zastopa skoraj izključno agregat G. lacustris, le po robu semkaj segata še agregata balcanicus in komareki. Domnevamo, da so takšni poselitvi prispevala izumiranja gamaridov v izjemno sušnih obdo-
62
Acta Biologica Slovenica, 62 (2), 2019
bjih. Tem je sledilo poseljevanje z evrieko vrsto G. lacustris, ki se morda edina razširja tudi s ptiči, in se je diverzificirala po naselitvi v Srednji Aziji. Zdaj tako na območju najdemo med drugim tudi ekstremno troglomorfno vrsto tega agregata in zelo aberantno jezersko vrsto.
Acknowledgements
The BS research in Turkmenistan was conducted during the field mission to Koytendag State Nature Reserve of Turkmenistan in May/ June 2015. The mission formed part of the Royal Society for the Protection of Birds' programme on biodiversity conservation in Turkmenistan in partnership with the Ministry for Nature Protection of Turkmenistan through their joint Memorandum of Understanding.
The ZH research was supported by National Natural Science Foundation of China (NSFC-
References
31772417). The DS research was supported by the Russian Foundation for Basic Research (grant No. 09/10-04-98544).
The samples from Central Asia have been processed by Gregor Bračko in the laboratory of the Biotehniška fakulteta, Univerza v Ljubljani, Ljubljana. One of the authors (BS) has been enjoying the hospitality of the same institution and financial support of the Slovenian Research Agency (program No. P1-0184, Zoology and speleobiology). Cene Fišer took some samples in his Tajikistan journey along the line extreme West - Dushanbe - Vanch.
Wim Vader helped us to decipher some geographical data. We particularly thank all collectors, listed in the supplementary table. Without them, such study would be impossible.
Thanks to Fabio Stoch who turned our attention to structural faults in the first MS. Thanks to Matejka Grašič for final polishing of our clumsy language.
Aladin, N.V., Plotnikov, I.S., 1993. Large saline lakes of former USSR: a summary review. Hydrobi-ologia, 267, 1-12.
Alanyali, F.S., Afik L., Özata A., Kutlu, H.M., 2011. Investigation of mitochondrial DNA of some West Anatolian Gammarus species (Crustacea: Amphipoda) by PCR-RAPD techniques. Turkish Journal of Biochemistry, 36, 207-212.
Alther, R., Fiser, C., Altermatt, F., 2016. Description of a widely distributed but overlooked amphipod species in the European Alps. Zoological Journal of the Linnean society, 179, 751-766.
Baetov, R., 2005. Lake Issyk-Kul: experience and lessons learned brief (Lake Basin Management Initiative) [available at http://worldlakes.org/uploads/13_Lake_Issyk-Kul_27February2006.pdf].
Balabanova, Z.M., 1949. Himizm ozer vostocnogo sklona Urala i Zaural'skoj ravniny. Trudy Ural'skogo otdela VNIORH, 4, 10-20.
Barnard, J.L., Barnard, C.M., 1983. Freshwater Amphipoda of the World. Hayfield Associates, Mt. Vernon, Virginia.
Bazikalova, A., Birstein., J.A., Taliev, D.N., 1946. Osmoticheskoe davlenie polostnoj zhidkosti boko-plavov ozera Bajkal. Doklady AN SSSR, 53, 293-295.
Behning, A., 1940. O nekotoryh rakoobraznyh okrestnostej Bakuriani. Trudy biologiceskih stancij Narkomprosa Gruzinskoj SSSR, 1, 11-58.
Bekman, M.Y., 1954. Biologija Gammarus lacustris Sars pribajkal'skih vodoemov. Akademija nauk, Trudy bajkal'skoj limnologiceskoj stancii, 14, 263-311.
Bening, A.L., Popova, A.N., 1947. Materialy po hidrobiologii reki Zangi ot istoka do g. Erevana (Materials on Hydrobiology of river Zangi, from the East to the city of Yerevan). Transmission of Sevan Hydrobiological Station, 8, 5-76.
Birstein, J.A., 1933. Malacostraca der Kutais-Hoehlen am Rion (Transkaukasus, Georgien). Zooloo-gischer Anzeiger, 104, 143-156.
Birstein, J.A., 1935. Zwei neue Gammariden aus Westasien. Zoologischer Anzeiger, 109, 293-296.
Sket et al.: Genus Gammarus in Central Asia
63
Birstejn, J.A., 1945. Zametka o presnovodnyh vyssih rakoobraznyh Turkmenii i Irana. Ucenye zapiski MGU, biologija, 83, 151-164.
Brodskij, K.A., 1935. Materialy k poznaniju fauny bespozvonocnyh gornyh potokov Srednej Azii, 1. Reka Issyk. Trudy Sredneaziatskogo gossudarstvennog universiteta, serija VIII-a Zoologija, 15, 1-112.
Brown, W.M., George, M., Wilson, A.C., 1979. Rapid evolution of animal mitochondrial DNA. Proceedings of National Academy Science USA. Genetics, 76, 1967-1971.
Buschkiel, A.L., 1931. Salmonidenzucht in Mitteleuropa. Handbuch der Binnenfischerei Mitteleuropas, 4, 161-348.
Caves, J.K., Sjostrom, D.J., Mix, H.T., Winnick, M.J., Chamberlain, C.P., 2014. Aridification of Central Asia and uplift of the Altai and Hangay mountains, Mongolia: stable isotope evidence. American journal of science, 314, 1171-1201.
Chappuis, P.A., 1955. Crustacés d'eau douce cavernicoles. In: Mission Henri Coiffait au Liban (1951). Archives de Zoologie Expérimentale et Générale, 91, 533-536.
Chaudri, M.A., Ghauria, A., Mahoon, M.S., 1987. Aquatic fauna of Swat vailey, Pakistan. Part III. Crustacea. Biologia, 24, 178-198.
Chevreux, E., 1908. Crustaces amphipodes. Etudes sur la faune du Turkestan basees sur le materiaux recueillis par D.D. Pedaschenko. Travaux de la Societe Imperiale des naturalistes de Saint-Petersbourg, 37, 91-100.
Chibilev, A.A., 1993. Zhizn posvyashchennay pticam - Nikolaj Alekseevich Zarudnyj. In: V glub stepej. Ocherki ob estestvoispytatelyah Orenburgskogo kraya. UIF Nauka, Ekaterinburg, 1-120.
Derzavin, A.N., 1923. Malacostraca presnyh vod Kamcatki. Russkij hidrobiologiceskij zurnal, 2, 180-194.
Derzavin, A.N., 1938. Bokoplavy Nahicevanskoi ASSR. Trudy zoologiceskogo insituta, Baku, 8, 163-184.
Derzavin, A.N., 1939. Presnovodnye perakaridy Talysa. Trudy zoologiceskogo instituta, Baku, 10, 43-57.
Dexbach, N.K., 1952. Mormysh (Gammarus lacustris) v vodoemah srednego Urala i Zaural'ya (ra-sprostranenie, ehkologiya, ispol'zovanie). Trudy Vsesoyuznogo gidrobiologicheskogo obshchestva AN SSSR, 4, 187-199.
Fryer, G., 1953. The occurrence of Gammarus lacustris G. O. Sars, and Gammaruspulex (L.) in Malham tarn, and a note on their morphological differences. Naturalist, Hull 1953, 155-156.
Haempel, O., 1908. Uber die Fortpflanzung und kunstliche Zucht des gemeinen Flohkrebses (G. pulex and G. fluviatilis). Allgemeine Fishereizeitung, 33, 86-89.
Hou, Z., Sket, B., Li, S., 2014. Phylogenetic analyses of Gammaridae crustacean reveal different diversification patterns among sister lineages in the Tethyan region. Cladistics, 30, 352-365.
Hou, Z., Sket, B., 2016. A review of Gammaridae (Crustacea: Amphipoda): the family extent, its evolutionary history, and taxonomic redefinition of genera. Zoological journal of the Linnean Society, 176, 323-348.
Hou, Z., Fu, J., Li, S., 2007. A molecular phylogeny of the genus Gammarus (Crustacea: Amphipoda) based on mitochondrial and nuclear gene sequences. Molecular phylogenetics and evolution, 45, 596-611.
Hou, Z., Li S., Platvoet, D., 2004. Three new species of the genus Gammarus from tributaries of the Ili river, China (Crustacea, Amphipoda, Gammaridae). Revue Suisse de zoologie, 111, 257-284.
Hou, Z., Sket, B., Fiser, C., Li, S., 2011. Eocene habitat shift from saline to freshwater promoted Tethyan amphipod diversification. Proceedings of the National Academy of Sciences of the United States ofAmerica, 108, 14533-14538.
Karaman, S., 1934. VI. Beitrag zur Kenntnis jugoslawischer Süsswasseramphipoden. Zoologischer Anzeiger, 107, 325-333.
64
Acta Biologica Slovenica, 62 (2), 2019
Karaman, G.S., Pinkster, S., 1977. Freshwater Gammarus species from Europe, North Africa and adjacent regions of Asia (Crustacea-Amphipoda), Part II. Gammarus pulex-group and related species. Bijdragen tot de Dierkunde, 47, 1-97.
Karaman, G.S., Pinkster, S., 1987. Freshwater Gammarus species from Europe, North Africa and adjacent regions of Asia (Crustacea-Amphipoda), Part III. Gammarus balcanicus-group and related species. Bijdragen tot de Dierkunde, 57, 207-260.
Karaman, G.S., 1993. Crustacea Amphipoda di acqua dolce. In: Karaman, G. (ed.), Fauna d'Italia, Vol. 31. Edizione Calderini, Bologna, 1-337.
Karaman, G.S., 2017. New data of poorly known species Gammarus orientalis (S. Karaman, 1934) (family Gammaridae) from Asia Minor (Contribution to the knowledge of the Amphipoda 292). Ecologica Montenegrina, 7, 639-653.
Karaman, S., Karaman, G., 1959. Gammarus (Fluviogammarus) triacanthus Schaferna, argaeus Va-vra und roeselii Gevrais am Balkan. Zavod za ribarstvo na NR Makedonija, Izdanija 2, 183-211.
Kurenkov, I.I., Mednikov, B.M., 1959. Identification guide to the malacostracans of fresh and brackish waters of Kamchatka. Izvestia Tihookeanskogo Naucno-Issledovatel'skogo Instituta, 47, 76-85.
Kydyraliev, A.K., 1990. Birds of lakes and mountain rivers of Kirgizia. Il'm, Frunze.
Labay, V.S., 1998. Besplatnyj referat i disertacija po biologii na temu: Fauna vyssih rakov (Crustacea Malacostraca) presnyh i solonovatyh vod ostrova Sakhalin. VAK RF 03.00.08, Zoologija.
Lien, A.M., 0stbye, K., 0stbye, E., 1996. Life cycle and morphology of an epigean and a hypogean population of Gammarus lacustris G.O. Sars (Amphipoda) in South Norway. Karst Waters Institute special publication, 2, 97-100.
MacDonald, K.S., Yampolsky, L., Duffy, J.E., 2005. Molecular and morphological evolution of the amphipod radiation of Lake Baikal. Molecular Phylogenetics and Evolution, 35, 323-343.
Maddison, D.R., Maddison, W.P., 2000. MacClade 4: Analysis of phylogeny and character evolution. Sinauer Associates, Sunderland, MA.
Martynov, A.V., 1930a. K poznaniju fauny Amphipoda ozera Issyk-kul. Issykkul'skaja ekspedicija 1928 g. In: Materialy komissii ekspedicionnyh issledovanij. Serij Kirgizskaja, Vol. 11. Akademija nauk SSSR, Leningrad, 51-70.
Martynov, A.V. 1930b. Fauna Amphipoda Teleckogo ozera i ego proishozdenie. Izvestija Gosudar-stvennogo gidrobiologiceskogo instituta, 29, 123-128.
Martynov, A.V., 1935. K poznaniju Amphipoda tekucih vod Turkestana. Trudy zoologiceskogo instituta Akademii nauk SSSR, 2, 411-508.
Martynov, A.V., Bening, A.L., 1948. O nekotoryh bokoplavah Srednej Azii. Trudy Zoologiceskogo instituta Akademii nauk SSSR, 7, 171-176.
Meng, K., Hou, Z., Li, S., 2003. A new species from Xinjiang, China (Crustacea: Amphipoda: Gammaridae). Acta Zootaxonomica Sinica, 28, 621-628.
Rachalewski, M., Banha, F., Grabowski M., Anastacio, P.M., 2017. Ectozoochory as a possible vector enhancing the spread of an alien amphipod Crangonyxpseudogracilis. Hydrobiologia, 717, 109-117.
Monod, T., 1931. Crustacés de Syrie. In: Gruvel, A. (ed): Les États de Syrie. Société d'Edition Geographiques, Maritimes et Coloniales, Paris, 397-435.
Moskric, A., 2016. Postavitev filogenetskega ogrodja rodu slepih postranic Niphargus (Crustacea: Amphipoda) na podlagi molekulskih metod (The formation of a phylogenetic framework for the genus Niphargus (Crustacea: Amphipoda) using molecular methods). PhD thesis. University of Ljubljani, Biotechnical faculty, Ljubljana.
0kland, K.A., 0kland, J., 1985. Factor interaction influencing the distribution of the freshwater »shrimp« Gammarus. Oecologia, 6, 364-367.
Peck, S.B., 1975. Amphipod dispersal in the fure of aquatic mammals. Canadian field naturalist, 89, 181-182.
Pirozhnikov, P.L., 1932. Issledovanie i ispol'zovanie vodoemov Sibiri: kratkoe rukovodstvo dlya kraevedov i mestnyh rabotnikov. Sovetskaya Aziya, Moscow.
Sket et al.: Genus Gammarus in Central Asia
65
Rauschert, M., Arntz, W.E., 2015. Antarctic macrobenthos. Arntz & Räuchert Selbstverlag, Bremen.
Rosine, W., 1956. On the transport of the common amphipod, Hyalella azteca, in South Dakota by the Mallard duck. Proceedings of the South Dakota Academy of Sciences, 35, 203.
Sars, G.O., 1895. An account of the Crustacea of Norway with short descriptions and figures of all the species, Vol. l. Amphipoda. Albert Cammermeyers Forlag (Lars Swanstram), Christiania and Copenhagen.
Sars, G.O., 1901. On the crustacean fauna of Central Asia. Annuaire du Musee zoologique de l'Academie imperiale des sciences de St. Petersbourg, 6, 130-164.
Segerstrale, S.G., 1954. The freshwater amphipods, Gammarus pulex (L.) and Gammarus lacustris G. O. Sars, in Denmark and Fennoscandia - a contribution to the lateand post-glacial immigration history of the aquatic fauna of northern Europe. Commentationes Biologicae, 15, 1-91.
Sidorov, D., Hou, Z., Sket, B., 2018. Three new remarkable amphipod species (Crustacea: Gammaridae) from springs and subterranean waters of Central Asia. Zootaxa, 4444, 437-461.
Sket, B., Hou, Z., 2018. Family Gammaridae (Crustacea: Amphipoda), mainly its Echinogammarus clade in SW Europe. Further elucidation of its phylogeny and taxonomy. Acta biologica Slovenica, 61(1), 93-102.
Stamatakis, A., 2014. RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics, 30, 1312-1313.
Stoch, F., Ruffo, S., 2003. Amphipoda. Checklist e distribuzione della fauna Italiana. Ministero dell'am-biente e della tutela del territorio, Roma.
Stock, J.H., Mirzajani, A.R., Vonk, R., Naderi, S., Kiabi, B.H., 1998. Limnic and brackish water Am-phipoda (Crustacea) from Iran. Beaufortia, 48, 174-234.
Straskraba, M., 1967. Re-examination of the taxonomic status of Niphargus indicus Chilton (Amphi-poda, Gammaridae) and its zoogeographical relations. Proceedings of the symposium on Crustacea held at Ernakulam, 1, 126-132.
Swofford, D.L., 2002. PAUP*: Phylogenetic analysis using parsimony (and other methods). Sinauer Associates, Sunderland, MA.
Snitnikov, A.V., 1980. Ozera Tjan'-Sanja i ih istorija. Nauka, Leningrad.
Snitnikov, A.V., Smirnova, N.P., 1981. Klimatologija, gidrologija i gidrofizika ozer Vnutrennego Tjan'--Sanja i ih istorija (tendencii prirodnogo razvitija). Nauka, Leningrad.
Thompson, J.D., Gibson, T. J., Plewniak, F., Jeanmougin, F., Higgins, D.G., 1997. The Clustal_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Research, 25, 4876-4882.
Thompson, P., Pradhan, R., Inskipp, C. et. all., 2001. Threatened birds of Asia. The BirdLife International Red Data Book, Parts A, B. BirdLife International, Cambridge.
Ueno, M., 1934. Yale North-India Expedition. Report on Amphipod Crustacea of the genus Gammarus. Memoirs of the Connecticut Academy of Art and Sciences, 10, 63-75.
Vainio, J.K., Väinölä, R., 2003. Refugial races and postglacial colonization history of the freshwater amphipod Gammarus lacustris in Northern Europe. Biological Journal of the Linnean Society, 79, 523-542.
Väinölä, R., Vainio, J.K., Usjak, S., Alther, R. Fiser, C., Altermatt, F., Witt, J.D.S. 2017. Global phylo-geography of Gammarus lacustris G.O. Sars, 1864, the most widespread freshwater amphipod species. Biodiversity Journal, 8, 561.
Wolff, C., Plessen, B. A. Dudashvilli, S., Breitenbach, S.F.M., Hai Cheng, H.S.F.M., Edwards, L.R., Strecker, M.R., 2017. Precipitation evolution of Central Asia during the last 5000 years. The Ho-locene, 27, 142-154.
Wundsch, H.H., 1922. Beiträge zur Biologie von Gammarus pulex. Archive für Hydrobiologie, 13, 478-531.
Zamanpoore, M., Grabowski, M., Poeckl, M., Schiemer, F., 2011. Taxonomic review of freshwater Gammarus (Crustacea: Amphipoda) from Iran. Zootaxa, 3140: 1-14.
66
Acta Biologica Slovenica, 62 (2), 2019
List of supplements (web edition)
Supplement S1: List of molecularly analyzed samples of the Gammarus lacustris aggregate, adequate to the tree in Fig. 2. Top samples (from the tree top to first Tajikistan samples) are only few representatives of the morphotype lacustris from Europe, the rest are all available samples of the aggregate. Highlighted samples are from Central Asia s.l..
Supplement S2: List of Gammarus spp., described from Central Asia, but not studied molecularly. Nomenclature and taxonomic subdivision are taken from original papers; the recent geographic, administrative, and orthographic changes have been considered as much as possible. For practical reasons, the original naming of taxa has been retained.
Supplement S3: Distribution map of Gammarus spp. in Central Asia. Red asterisks with names -important cities. Reversed triangles - molecularly not defined Gammarus spp., letters indicate abbreviated taxonomic names (see Suppl. S2). Rings - molecularly defined taxa of agg. lacustris, numerals are OTU designations (as in the second column in Suppl. S1); pale ribbons connect localities of G. cf. subaequalis-Garlyk, hatched ribbon connection with Martynov's G. subaequalis and G. turanus. Squares - agg. balcanicus; diamond (G. turcomanicus) - agg. komareki. Adequate to Figure 1 in the paper.
Supplement S4: Additional taxonomical clarification. (1) Type population of Gammarus lacustris Sars, 1863. (2) Identity of 'G.balcanicus 'fromAlma-Ata. (3) Corrected classification of 'Anisogammarus' madyensis Chaudri, Ghauri & Mahoon, 1987.
Supplement / Dodatek
Supplement SI: List of molecularly analyzed samples of the Gammanis lacustris aggregate, corresponding to the tree in Fig. 2. Top samples (from the tree top to first Tajikistan samples) are only few representatives of the morphotype lacustris from Europe, the rest are all available samples of the aggregate. Highlighted samples are from Central Asia s.l..
Dodatek SI: Popis molekulsko analiziranih vzorcev agregata Gammanis lacustris, ki se ujema s sliko 2. Vrhnji vzorci (od vrha drevesa do prvih tadžikistanskih vzorcev) so le redki predstavniki morfotipa lacustris iz Evrope, preostalo pa so vsi dosegljivi vzorci agregata. Zasenčeni (obarvani) vzorci so iz Centralne Azije s.l..
Taxon - working names; (Type - from type locality)/ Takson - delovna imena (Type - iz tipske lokalitete)	Designation of OTUs in the tree/ Oznaka OUT v drevesu	Locality (r. = stream; c. = cave; N, S, W, E = north, south, west, east)/ Lokaliteta (r. = potok; c. = jama; N, S, W, E = sever, jug, zahod, vzhod.)	Collector(s) or Reference (with year)/ Zbiralec ali vir podatka (z letnico)	Habitat/ Habitat	Country/ Drzava	Voucher number/ Številka vaučerja	COI	28S
G. alpinus Alther, Fišer & Altermatt 2016	GalpIND (7 OTUs)	Alps	Alther et al. 2016	lakes	Austria S (Europe)		KX283210	
G. cf. lacustris	OOlGlacBle	lake Blejsko jezero. Bled	B. Sket	lake	Slovenia NW (Europe)		JF965915	JF965728
G. alpinus	GalpIND (50 OTUs)	Alps	Alther et al. 2016	lakes	Austria N (Europe)		KX283184	
G. alpinus	GalpIND (6 OTUs)	Alps	Alther et al. 2016	lakes	Austria N (Europe)		KX283188	
G. alpinus	GalpIND (9 OTUs)	Alps	Alther et al. 2016	lakes	France (Europe)		KX283195	
G. lacustris G.O. Sars 1864 (close to Type)	473Glacust	lake Birisjoen, Vâgâ, Oppland	J.Vainio	lake	Norway S (Europe)	SLOCHN473	MK066420	MK066380
G. lacustris	KJ831288	lake Ulvenvann (?Oslo?)	K.Ostbye et al.	lake	Norway S (Europe)		KJ831288	
G. lacustris	KJ831277	lake Tinho, Eidfjord, Hordaland (Bergen)	K.Ostbye	lake	Norway SW (Europe)		KJ831277	
G. lacustris	KJ831281	Sandagrotta Cave, Skrim, Kongsberg S	K.Ostbye	cave	Norway S (Europe)		KJ831281	
G. lacustris	KJ831279	Sandagrotta Cave, Skrim, Kongsberg S	K.Ostbye	cave	Norway S (Europe)		KJ831279	
G. alpinus	GalpIND064	GalpIND064	Alther et al. 2016	lake	Alps (Europe)		KX283190	
G. lacustris	KJ831278	Sandagrotta	K.Ostbye	cave	Norway S (Europe)		KJ831278	
G. lacustris	KJ831280	Sandagrrotta	K.Ostbye	cave	Norway S (Europe)		KJ831280	
G. lacustris	KJ831287	lake Ulvenvann	K.Ostbye	lake	Norway S (Europe)		KJ831287	
G. lacustris	377Glacustr	lake Tinho, Eidfjord, Hordaland	K.Ostbye	lake	Norway S (Europe)	SLOCHN377	MK066421	MK066381
G. cf. kischineffensis	lOlGspUDon	r. Severski Donee, Biškin, Veliki Burluk, Kharkiv	P. Trontelj	river	Ukraine (Europe)		JF965910	JF965722
G. sp	262 GGumas	Gorno Badakhashan, SE of Gumas	C.Fišer	spring	Tajikistan		JF965966	JF965796
G. cf. bergi	370Issy	lake Ysyk Kol, littoral, Cholpon Ata	B. Sket	lake	Kyrgyzstan	SLOCHN370	MK066422	MK066382
G. bergi Martynov 1930 (close to Type)	361GIsy IK	lake Ysyk Kol, ittoral, Barskoon, NW Kichi Jargylchak	B. Sket	lake	Kyrgyzstan	SLOCHN361	MK066423	MK066383
G. bergi	366Gissy	Lake Ysyk Kol, -30m, 12 km E Kadzhi Say, Bokonbaevo	B. Sket	lake	Kyrgyzstan	SLOCHN366	MK066424	MK066384
G. bergi	357Gbergi	Lake Ysyk Kol littoral, Bokonbaevo, 12 km E Kadzhi Say	B. Sket	lake, -30m	Kyrgyzstan	SLOCHN357	MK066425	MK066385
G.bergi	367Gbergi	Lake Ysyk Kol, -10m, Cholpon Ata KG,	B. Sket	lake, -10m	Kyrgyzstan	SLOCHN367	MK066426	MK066386
G. cf. lacustris	KJ831286	lake Lille Lauarvann, Kongsberg	K.Ostbye	lake	Norway S (Europe)		KJ831286	
G. cf. lacustris	KJ831282	lake Lille Lauarvann, Kongsberg	K.Ostbye	lake	Norway S (Europe)		KJ831282	
G. cf. lacustris	KJ831285	lake Lille Lauarvann, Kongsberg	K.Ostbye	lake	Norway S (Europe)		KJ831285	
G. cf. lacustris	KJ831283	lake Lille Lauarvann, Kongsberg	K.Ostbye	lake	Norway S (Europe)		KJ831283	
G. cf. lacustris	KJ831284	lake Lille Lauarvann, Kongsberg	K.Ostbye	lake	Norway S (Europe)		KJ831284	
G. cf. lacustris	GlacusNACA		K.Ostbye		Canada		AY529052	AY529073
G. cf. lacustris	GlacusWAUSA				USA		AY926673	
G. cf. lacustris	GlacusVICA				Canada		AY926672	
								
1
G. cf subaequalis-Garlyk	TAD3	Zeravshan valley	C.Fiser	spring	Tajikistan		JF965968	JF965798
G. cf subaequalis-Garlyk	465Gpamir	lake Karakul, Pamir	D. Kryvokliyzha	lake	Tajikistan NE	SLOCHN465	MK066427	MK066387
G. cf subaequalis-Garlyk	423GIskand	lake Iskandarkul	C.Fiser	lake	Tajikistan	SLOCHN423	MK066428	MK066388
G. cf subaequalis-Garlyk ("Type")	516Garlyk	Kara Bulak, Karlyuk, Koytendag	B.Sket	stream	Turkmenistan E		MG655630	MK066389
G. cf subaequalis-Garlyk	369Gissy ,	lake Ysyk Kol, -10m, Cholpon Ata,	B.Sket	lake -10m	Kyrgyzstan	SLOCHN369	MK066429	MK066390
G. cf subaequalis-Garlyk	261GAalsud	lake Alaudin	C.Fiser	lake	Tajikistan		JF965965	JF965795
G. cf subaequalis-Garlyk	S135	Pamir, GBAO, Ishkoshim distr. vil. Zumudg	D.Palatov	stream	Tajikistan	S135	MK160078	
G. cf subaequalis-Garlyk	TAD1	Khujand, Sogdiyskaja obi.	C.Fiser	spring	Tajikistan		JF965967	JF965797
G. cf alpinus	GlacIND048(-056), 8 OTUs	Alps	Alther et al. 2016	lake	Alps (Europe)		KX283247	
G. sp.	372 Gkadam	Kyzyl Bulak, Kadamzhai	B.Sket	stream	Kyrgyzstan -SW	SLOCHN372	MK066430	MK066391
G. sp.	467 Gpamir	lake Bulunkul, Pamir	S. Litvinchuk	lake	Tajikistan SE	SLOCHN467	MK066431	MK066392
G. cf. lacustris	Glacus423	Altun, Xinjiang	Hou et al 2007	lake	China	IZCASIA423	EF570321	
G. cf. lacustris	Glacus 618	618 Xiainza, Tibet (?)	Hou et al 2011		China Tibet		EF570322	
G. cf. lacustris	275Granwu	lake Ranwu, Yale	A.Brancelj		China Tibet	SLOCHN275	MK066432	MK066393
G. cf. lacustris	GlacusHVMongolia	? lake Hovsgol	Macdonald et al 2005		Mongolia		AY926670	
G. decorosus Meng, Hou & Li 2003	Gdecoro436	Uramqi	Hou et al 2011		China Xinjiang		EF570303	EF5 82949
G. cf. lacustris	1104Glacul	Alxa Zuoqi, Inner Mongolia	Hou et al 2014		China Inner Mongolia		JF965918	JF965733
G. cf. lacustris	1136Glacul	Selenga	Hou et al 2011	river	Mongolia		JF965917	JF965732
G. cf. lacustris	Glacustris BK	Olhon Island, lake Bajkal	??? V.Tahteev	lake	Russia Bajkal		AY926671	
G. cf. lacustris	859 Glacus 1	Haidong, Xunliua, Qinghai	D. Qi		China Qinghai	IZCASIA0859	MK066433	
G. cf. lacustris	Glacust 326	Fangshan	Hou et al 2007		China Beijing		EF570318	
G. cf. lacustris	Glacust 323	Fangshan	Hou et al 2007		China Beijing		EF570317	EF582964
G. cf. lacustris	Glacust292	Beijing, Yanqing	Hou et al 2007		China		EF570315	
G. cf. lacustris	696Glacust	Changji, Uramqi,	CHN	CHN 696 (with 436)	China Xinjiang		JF965916	JF965730
G. cf. lacustris	498Gladak	Leh, Ladakli	G.Mazepa		India J&K	SLOCHN498	MK066434	MK066395
G. cf. lacustris	478Glacust	lake Shira	K. P. Veresshagina, E.S.Zadereev, A.G.Degermebdzhy	lake	Russia Khakassia	SLOCHN478	MK066435	MK066394
G. cf. lacustris	S133	Khovd aimag, Chono-Kharaikliin-gol river	D.Palatov	stream	Mongolia	S133	MK160077	
G. cf. lacustris	Glacust340	Hoh Xil	CHN		China Qinghai		EF570320	EF582965
G. cf. lacustris	Glacust624	Qinghai	Hou et al 2007		China		EF570323	
G. cf. lacustris	342Gladakli	Ladakli	G.Mazepa		India J&K	SLOCHN342	MK066436	
G. sp3	374Gkadam	Abshir Ata, Abshirsai, Khodzhaaryk	B.Sket,	spring	Kyrgyzstan SW	SLOCHN374	MK066437	MK066396
G. sp3	375Gabshirs	Abshir Ata-waterfall, Abshirsai, Khodzhaaryk	B.Sket,	spring-waterfall	Kyrgyzstan SW	SLOCHN375	MK066438	MK066397
								
G. sp4	TAD4	oasis Chilychor, Shaartuz distr.	Hou etal 2011	spring	Tajikistan SW		JF965970	JF965800
G. sp4	453Chilych	oasis Chilychor, Shartuuz d., Khatlonskaya oblast	G.Mazepa		Tajikistan SW	SLOCHN453	MK066439	MK066398
G. cf. italicus	075GitalIT	Lazio, Rieti, Terme di Cotilia	C.Fiser	spring	Italia (Europe)		JF965904	JF965716
G. cf. pseudosyriacus	286Gafqa	El Mnaitra, Qartaba	B.Sket,	spring	Libanon	SLOCHN286	MK066440	MK066415
G. cf. italicus	240GitalLB	spring Labour	L.Simon	spring	Libanon		JF965907	JF965719
G. cf. pseudosyriacus	499Gquarta	El Aaqoura, Qartaba	B.Sket,	spring	Libanon	SLOCHN499	MK066441	MK066399
G. cf. lobifer	345Globifer	Yasooj, Chaharmah	V.Akmali (det. Akmali &BS)		Iran	SLOCHN345	MK066442	MK066416
2
G. elvirae Iamiilli & Ruffo 2002 (Type)	442Gelvirae	Fontana Tiburto, Ciciliano, Lazio	l&d F. Stoch	spring	Italy (Europe)	SLOCHN442	MK066443	MK066400
G. cf. acalceolatus	168GGoigou	Timahoite, Goigou	R.Verovnik		Morocco (Africa)		JF965831	JF965636
G. sp 1	198 GTalas	river Talas, Dzhambul oblast", Taraz	G.Mazepa	stream	Kazakhstan		JF965963	JF965791
G. sp2 (short antemiae)	363Gtokni	Tokniok, Bi§kek	B.Sket	spring	Kyrgyzstan	SLOCHN363	MK066448	MK066402
G. sp2	376Gtyup	4kni NW Tyup, Karakol	B.Sket	spring	Kyrgyzstan	SLOCHN376	MK066449	MK066403
G. sp2	358Gkara	5km NW Karakol	B.Sket	spring	Kyrgyzstan	SLOCHN358	MK066450	MK066404
G. sp2	368Gissy	lake Ysyk Kol, -10m, Cholpon Ata	B.Sket	lake	Kyrgyzstan	SLOCHN368	MK066451	MK066405
G. sp2	365Gissy	lake Ysyk Kol, -70m, Cholpon Ata	B.Sket	lake, -70m	Kyrgyzstan	SLOCHN365	MK066452	MK066406
G. sp2	364Gissy	lake Ysyk Kol, -70m, Cholpon Ata	B.Sket	lake, -70m	Kyrgyzstan	SLOCHN364	MK066453	MK066407
G. sp2	362Gka	10km NE Karakol	B.Sket	spring	Kyrgyzstan	SLOCHN362	MK066454	MK066401
G. decorosus Meng, Hou & Li 2003	693GGdecoro	LIrumqi, Xinjiang	CHN		China Xinjiang		JF965875	JF965684
G. troglomorphus Sidorov, Hou & Sket 2018 (Type)	513Gtroglm	Suw Oyuk, Karlyuk	B.Sket	groundwater	Turkmenistan	SLOCHN513	MG655628	MK066408
G. troglomorphus	N2TLIM	Suw Oyuk, Karlyuk	B.Sket	groundwater	Turkmenistan	SLOCHN513a	MK066444	
G. parvioculatus Sidorov, Hou & Sket 2018 (Type)	515Gparvoc	Hojeipil, Koytendag	B.Sket	spring	Turkmenistan	SLOCHN515	MG655629	MK066409
G. varsoviensis Jazdzewski, 1975 (Type)	174varsov	Secymin village	M. Grabowski		Poland (Europe)		JF965987	JF965818
G. cf. varsoviensis	420Gniykol	Povd. Bug, Pervomaysk d., Mykolayiv Reg.	S.Sidorovsky	stream	LTkraine (Europe)	SLOCHN420	MK066445	MK066418
G. sp.	TAD2	Zeravshan, Sogdiyskaja oblast	L. G. Mazepa	spring	Tajikistan		JF965969	JF965799
G. cf. komareki	454Gzeravsh	kishlak Veshkand, Sugd	L.G.Mazepa		Tajikistan	SLOCHN454	MK066455	MK066410
G. tianshan Zhao, Meng & Hou 2017(?Type)	432 G. tianshan	Zhaosu, Xinjiang CHN 432	CHN		China		EF570327	EF582971
G. tianshan	371 Gkara	371 Karakol	CHN		Kyrgyzstan	SLOCHN371	MK066456	MK066412
G. sp.	360 bish	Ala Archa, Bi§kek	B.Sket	spring	Kyrgyzstan	SLOCHN3ÔO	MK066446	MK066411
G. sp.	257Gteleck	lake Teleckoe ozero	M.Schletterer	lake	Russia		JF965964	JF965792
G. cf. lobifer	235Globif	Khuzestan Prov., Choram	A.Sari & B.Sket	spring	Iran		JF965920	JF965735
G. sp.	343Gkashm	Kashmir ?	G.Mazepa		India J&K	SLOCHN343	MK066457	MK066413
G. cf inberbus (?)	236Gimberb	Kohgiluyeh prov., S Yasuj	A.Sari & B.Sket	spring	Iran		JF965902	JF965714
G. cf. zagrosensis	351Gzagros	Sepidan, Fars	V.Akmali (det. Akmali & BS)		Iran	SLOCHN351	MK066458	MK066417
G. italicus Goedmakers & Pinkster 1977 (Type)	161Gitalic	Sassari (Sardegna),	F. Stoch		Italia (Europe)		JF965905	JF965717
G. italicus	444Gitalic	Badde Selo, Sardegna	F. Stoch		Italia (Europe)	SLOCHN444	MK066447	MK066414
G cf. crinicaudatus Stock,	234Gcrinic	Bakhtiari Prov., Lordegan	A.Sari & B.Sket	spring	Iran		JF965874	JF965683
G. cf. pseudosyriacus	233Gpsdsy	Bakhtiari Prov., Lordegan	A.Sari & B.Sket	spring	Iran		JF965937	JF965759
OUTGROUP/ZUNANJIKI								
G. cf. uludagi	116Gulud	island Evia	G.LTrbanič		Greece (Europe)		JF965986	
G. pulex Linne 1758	022GpulUpp	r. Haga, Kvarnbo, LIppsala	D. Andersson		Sweden (Europe)		JF965939	
G. bosniacus Schäfema 1922 (Type)	137Gbosni	Vrela Bosne, Ilidža, Sarajevo	B.Sket	spring-lake	Bosnia & Herzegovina (Europe)		JF965872	
G. cf. balcanicus	018GbalZen	Zenica	T. Rada	spring	Bosnia & Herzegovina (Europe)		JF965837	
G. pavlovici S.Karaman 1931 (Type)	084GpavMA	Rašče, Skopje	B.Sket	spring-waterwork	Macedonia (Europe)		JF965930	
	563Gtastiensis						EF570347	EF582993
Gammaras Tturcomanicus	512Gturcom		B.Sket	springs	Turkmenistan	SLOCHN512	MK066459	MK066419
	269Gkomar						JF965913	JF965725
	104Gfoss						JF965886	JF965696
3
Supplement S2: List of Gcmimarus spp., described from Central Asia, but not studied molecularly. Nomenclature and taxonomic subdivision are taken from original papers; the recent geographic, administrative, and orthographic changes have been considered as much as possible. For practical reasons, the original naming of taxa has been retained.
Dodatek S2: Popis vrst rodu Gammanis, opisanih iz Centralne Azije, vendar ne molekulsko analiziranih. Nomenklatura in taksonomska delitev so vzeti iz izvirnih del; sodobne geografske, administrativne in pravopisne spremembe smo upoštevali, kolikor je bilo mogoče. Iz praktičnih razlogov smo ohranili izvorno poimenovanje.
Aggregate lacustris
The majority of following taxa or populations are morphologically within the diversity of the aggregate lacustris.
(1)	Rivnlogammanis turamis Martynov, 1935 with subordinate taxa:
R. t. turamis Martynov, 1935, Uzbekistan, Tashkent province, Chimgan, river Chimganka, near the sanatorium;
R. t. morpha subnivalis Martynov, 1935, Uzbekistan, Tashkent province, source of the river Chimganka;
R. t. natio kuihikensis Martynov, 1935, Uzbekistan, 9 km from Tashkent, spring at town Kuylyuk;
R. t. subsp. coxalis Martynov, 1935, Kazakhstan SE, range Alexandrovskij, Taraz (= Aulie-Ata), gorge Kulan, near Dmitrievka;
R. t. forma karabasicus Martynov, 1935, Kazakhstan S,E Karatau mountains, torrent Karabas-tau;
R. t. karabasicus morpha excisus Martynov, 1935, Kazakhstan S, E Karatau, at the mount Bahai--bas-tau near Galkino.
(2)	Rivnlogammanis subaequalis Martynov, 1935,
with subordinate taxa (and/or synonyma):
R. s. subaequalis Martynov, 1935, Kazakhstan S, eastern Karatau, limnocrene at Galkino (= Zertas);
R. s. forma zanidnyi Martynov, 1935, Kazakhstan S, in Bulaki, Karatau, (position insecure);
R. s. forma bianchii Martynov, 1935, Kazakhstan S, Taraz (= Aulie-ata), close to Talas;
R. s. bianchii morpha compressus Kazakhstan S, Jambyl Province, spring at Taraz (= Aulie-ata) and surroundings;
Gammarus (Rivnlogammanis) svriacus subaequalis (Martynov 1935), Tajikistan, springs at fortress Hisor (= Gissar, Hissar), r. Hanakinka, kishlak near st. Hanaka, kislak Hodzambio; 'very common, widely spread in Tajikistan' (Birstejn, 1948)
(3)	Gammarus korbuensis Martynov 1930.
Distribution: Russia, Altaj, river Korbu, affluent of Teleckoe ozero (Altyn Kol).
With Gammarus korbuensis forma reducta Martynov 1930, Russia, Teleckoe ozero (Altyn Kol), Altaj.
(4)	Gammarus teletzkensis Martynov 1930
(syn. Gammaruspulex G. O. Sars 1903)
Distribution: Russia, Altaj, Teleckoe ozero (Altyn Kol),
(5)	Gammarus ocellatus Martynov, 1930 with subordinate taxa
G. o. ocellatus Martynov, 1930
Distribution: Kyrgyzstan, Ysyk Kol (Issyk-Kul Lake; Martynov, 1930a); Tajikistan, Iskander-kul Lake, mouth of the Serima river (Iskander-kul Lake basin) and small Verin-kul lake (unknown locality) (Martynov, Behning, 1948).
G. o. subsp. angulatus Martynov 1930, Teleckoe ozero, Altaj, Russia.
(6)	Gammarus inberbus Karaman & Pinkster, 1977
Distribution: Kyrgyzstan, Koysara (Koisara; and some undefined places in) Ysyk Kol (Issyk-Kul Lake; Karaman & Pinkster, 1977).
All our gammarids from SE part of the lake are provided with slight dorsal elevations on urosomites, while in G. inberbus pleonites and urosomites are explicitely 'smooth, without dorsal elevations or excavations' (Karaman & Pinkster, 1977).
(7)Rivnlogammanis	truncatus Martynov, 1935
with subordinate taxa:
R. t. truncatus Martynov, 1935, Uzbekistan S (eastern Buchara, localization not reliable), village Sangirdak, r. Surkhan (Surkhanderya?) (? S) (Synonym Gammarus sanginlaki G. Karaman, 1984).
R. t. (Mart.) sbsp. montanus Martynov 1935 (homonym, nec G. pavlovici montanus S. Karaman 1929, nec Gammarus montanus Hou, Li & Platvoet, 2004; synonym Gammarus sangirdaki hissari G. Karaman, 1984.), Tajikistan, locality Han-tahta, northern Hissar range.
(8)	Gammarus gracilis Martynov, 1935, Kazakhstan, eastern Karatau, near Shymkent (= Tshimkent). Syn. Gammarus chimkenti G. Karaman, 1984.
According to the author, this is a separate species, but morphologically and geographically very close to G. turamis.
(9)	Gammarus angusticoxalis Martynov, 1935
Kyrgyzstan, near Bishkek (= Pishpek), rivers Dzhumgan and Ala-archa (Chu river basin?) (Martynov, 1935). Uropod III endopodite less than 50% exopodite length.
(10)	Gammarus montaniformis Sidorov, 2012
Kazakhstan, eastern Tien-Shan, Dzungarian Alatau, ca 2200 m a.s.l. (Sidorov, 2012).
Characterised by comparatively short uropod III endopodite.
(11)	Rivulogammarus brevicomis Martynov, 1935,
Kazakhstan S, Karatau, springs at rivulets Kichkin, Ishfan-Mazar.
Martynov: 'A separate species, but close to R. turamis'.
(12)	Gammarus alius Sidorov, 2012
Distribution: Kyrgyzstan, lake Chatyr-kul, 3555 m a.s.l.
Its unique character is a pair of hooks on the epistome of the upper lip.
Aberrant taxa of uncertain position
(13) Rivulogammarus hirsutus Martynov, 193 5 ( syn. : Rivulogammarus brachyurus Birstein, 1935, Gammarus birsteini Karaman et Pinkster, 1977) with subordinate taxa:
R. h. hirsutus Martynov, 1935, S Kazakhstan, NE Karatau; Syr Darya Mountains, Talas-Alatau (Birstejn 1935);
R. h. var. oligochaetus Birstejn, 1935, S Kazakhstan, Syr-Darya Mountains; Karatau Mountains; near Buchujtau;
R. h. morpha hirsiitissimus Martynov 1935, Kazakhstan, Kel'ty-Mashat, eastern Karatau.
(14) Issykogammants hamatus Chevreux 1908
Distribution: Kyrgyzstan NE, only in the lake Ysyk Kol, from different depths.
It was described as a separate genus. But if we only overlook the recast of the pereopod coxae, the species fits easily into the diversity of the aggregate lacustris. Unfortunately, a molecular analysis was impossible and the oldfashioned presentation of the morphology, even done mainly on the female, does not allow a serious réévaluation of the species' taxonomy and phytogeny.
Aggregate balcanicus
Members of the agg. balcanicus are comparatively easily recognizable morphologicaly. The reliability of such an identification is however tow, but in our case it does not contradict the few molecularly supported data. Characteristic is the poor setation of pereopods III-IV and of uropod III. In the latter, the spines along the exopodite outer margin are joined by only few and short finer setae.
(15)	Rhmlogammants spinulatus Martynov, 1935
Distribution: Kazakhstan, mountain river (torrent) Issyk in Semireche (= Zhetisu) district (E of Almaty).
(16)	Gammantspellucidus Guijanova, 1930.
Distribution: Russia, Krasnojarsk, river Enisej, Ladejskij perekat.
(17)	Gammanis angustatiis Martynov 1930 with subordinate taxa
G. a. angustatiis Martynov 1930, Russia, upper Ob, torrent Injuska, affluent of Ini; Kyrgizstan NE, river Alamedinka near Bi§kek.
G. a. forma obensis Martynov 1930, Russia, Novosibirsk, river Ob.
(17a) Gammanis bivvipodus Hou, Li & Platvoet, 2004 from China, Xinjiang, Xinyuan, W ofUrumqi.
(17b) Gammanis 'balcanicus' (Gammanis (Rivulogammants) balcanicus Birstejn, 1950) from Russia, Oblast' Kemerovo, Jurginskij rajon)
Aggregate pulex
(18)	The only seemingly reliable indications on G. pulex are by Birstejn 1945 from the rivers Sherlok and Firjuzinka near A§gabat (Ashgabat) (central Kopetdag, neigbour to G. turcomanicus).
(19)	Rivulogammants pulex forma mesasiaticus Martynov 1936 from Kyrgyzstan NE. Two 'sub-forms', one in springs and in river, another in stagnant waters (ponds) near Tokmok and Bi§kek (= Frunze) do not exhibit any characters specific to G. pulex. Our samples from surroundings of Tokmok and Bi§kek contain only members of the agg. lacustris.
5
Supplement S3: Distribution map of Gammarus spp. in Central Asia. Red asterisks with names - important cities. Reversed triangles - molecularly not defined Gammarus spp., letters indicate abbreviated taxonomic names (see Suppl. S2). Rings - molecularly defined taxa of agg. lacustris, numerals are OTU designations (as in the second column in Suppl. SI); pale ribbons connect localities of G. cf. siibaequalis-Garlyk, hatched ribbon connection with Martynov's G. siibaequalis and G. turanus. Squares - agg. balcanicus; diamond (G. tiircomanicus) - agg. komareki, upright triangles - Sarothrogammanis group. Corresponding to Figure 1 in the paper.
Dodatek S3: Karta razširjenosti Gammarus spp. v Centralni Aziji. Rdeče zvezdice z imeni - pomembni kraji. Obrnjeni trikotniki - molekulsko nedefinirane Gammarus spp., črke so okrajšana taksonomska imena (glej Dodatek S2). Krožci - molekulsko definirana taksa agregata lacustris, številke so oznake OTUjev (kot v drugem stolpcu v Dodatku SI); svetle proge povezujejo lokalitete tipa G. cf. siibaequalis-Garlyk, prekinjena proga povezavo z G. siibaequalis in G. turamis Martynova. Kvadratki - agregat balcanicus; romb (G. tiircomanicus) - agregat komarekr, pokončni trikotniki - skupina Sarothrogammanis. Se ujema s Sliko 1 v članku.
6
Supplement S4: Additional taxonomical clarification Dodatek S4: Dodatna taksonomska pojasnila.
(1)	Type population of Gammarus lacustris Sars, 1863.
Gammanis lacustris the name-bearing species of the aggregate lacustris and topotype population of it, was described by Sars without defining the type locality. Karaman & Pinkster (1977) selected 'Norway, Selsvand, Vage' as 'restricted type-locality'. Since we could not detect such a locality, W. Vader (pers. com.) helped us with identification of Selsvand as Selsvatn(et) in the community of Vaga, province of Oppland in mid Norway. This is the correct spelling of the selected type locality. Our sample 473 is from the lake Birisjoen, Vaga Municipality, ca. 200 kmNNW Oslo and only ca 35. km SWfrom Selsvatnet, hydrographi-cally in the same sytem; we consider the sample 473 as belonging to the topotype (metapopulation of Gammanis lacustris Sars, 1863.
(2)	Identity of 'G.balcanicus' from Alma-Ata.
The origin of G. balcanicus of MacDonald et al., 2005, reportedly from Alma-Ata, after a joint reconsideration with some authors, could not be localized with certainly. Molecularly, this OTU is related to populations from Krym (Ukraine). On the other hand, it is very likely that R. spinulatus from the same area (E of Almaty) really belongs to the agg. balcanicus, as much as morphological characters are reliable.
(3)	Corrected classification of 'Aiiisogaimnarus'madyensis Chaudri, Ghauri & Mahoon, 1987
The morphology of 'Anisogammarus' madyensis clearly positions the species into the Sarothrogam-
marus group of genera, more precisely the genus Sarothrogammanis. We will provisionally classify it as Sarothrogammarus madyensis (Chaudri, Ghauri & Mahoon, 1987). Its provenience in the Swat Valley is not in contrast with the group biogeography, it only slightly extends its area and adds Pakistan. On the other hand, Anisogammaridae are alien to the area of W Asia.
7