Origin and recent endemic divergence of a Caspian Mysis species flock with affinities to the "glacial relict" crustaceans in boreal lakes.
Received July 26, 1993. Accepted July 6, 1994.
The endemic complexes of closely related species that characterize the faunas of ancient lakes present intriguing challenges to evolutionary research (e.g., Brooks 1950; Echelle and Kornfield 1984; Fryer 1991). Among the principal problems are the mechanism of speciation in a confined lacustrine space, and the avoidance of competitive exclusion among closely related sympatric taxa. For any particular complex, or species flock, the age of divergence and the rate of morphological differentiation, and the number and zoogeographical origins of the founder taxa are often controversial. The most prominent instances of endemic lacustrine diversification are found in the cichlid fishes of several East African lakes (e.g., Fryer and Iles 1972) and the gammarid amphipod crustaceans of Lake Baikal (Kozhov 1963). More generally, these and other similarly old basins harbor species flocks in several animal groups, with a degree of diversity varying from group to group.
Molecular approaches are shedding new light on the dynamics of intralacustrine species flock evolution. Close monophyletic relationships and a rapid morphological diversification in the species-rich Lake Victoria haplochromine cichlid complex have been inferred from allozyme and mtDNA data (Sage et al. 1984; Meyer et al. 1990). On the other hand, studies of Lake Tanganyika cichlids indicate that, under intrabasin geographical isolation, the fish may also retain a stable morphology over long periods of time (Sturmbauer and Meyer 1992), and that the total cichlid diversity of a lake may represent several lineages older than the lake itself (Nishida 1991). The modes and rates of speciation and of morphological divergence vary greatly even among closely related groups.
In an assessment of allozyme differentiation in a small endemic Mysis species assemblage in the Caspian Sea, this paper extends the molecular approach to the study of intralacustrine species flock evolution from vertebrates to invertebrates (mysidacean crustaceans), from the nearshore benthic to the deepwater pelagic environment, and geographically to another ancient basin, illustrating both parallels and contrasts to the fish flock data. In a comparison of congeneric mysid species groups from three different zoogeographical zones--the Caspian Sea, boreal freshwater lakes, and northern marine waters--the results suggest widely different relative rates of morphological and molecular divergence in different environments, supplying another example of rapid intralacustrine diversification contrasted by a morphological stasis in allopatry.
In addition to aspects of species flock evolution, another main objective of the study is to evaluate phylogenetic and zoogeographical hypotheses related to the origin of the Arctic element in the Caspian fauna and, particularly, the relationships of this element to the "glacial-relict" element in young boreal lakes. The background to this controversial issue is reviewed below.
Zoogeographical Settings: The Arctic Element in the Caspian Fauna
The fauna of the enclosed, brackish Caspian Sea consists mainly of autochthonous species and genera that are thought to have evolved in the Ponto-Caspian basin independently since the early Pliocene. About half of the species are currently strictly endemic to the Caspian Sea. Although largely of marine ancestry, the fauna has parallels with those of ancient freshwater lakes. It comprises several moderate-sized species flocks (endemic genera), for example, in fishes and various groups of mollusks and crustaceans (Zenkevitch 1963; Mordukhai-Boltovskoi 1964; Birstein et al. 1968).
Apart from the prevailing autochthonous temperate comportent, the Caspian Sea harbors a distinct zoogeographical assemblage referred to as the Arctic element, dominating the deeper parts of the basin (50-1000 m; salinity 12-13%). The element includes crustaceans from diverse genera (Monoporeia, Gammaracanthus, and Onisimus [= Pseudalibrotus] [Amphipoda], Mysis [Mysidacea], Saduria [Isopoda], and Limnocalanus [Copepoda]), which are all also represented by closely related forms in coastal and estuarine waters of the Arctic basin (e.g., Hogbom 1917; Zenkevitch 1963). Most of the genera have single species in the Caspian, but Onisimus has two, and Mysis comprises a flock of four endemic species. Although the endemic diversity in Caspian Mysis falls far short of that in the more spectacular lacustrine assemblages, it fits Ribbink's (1984) definition of a species flock, in which the essential feature is a disproportionate local diversity as compared to that in related allopatric forms, rather than the absolute species number itself.
Most of the Arctic genera also inhabit a third geographically and ecologically distinct zone--the young boreal and subarctic lakes in the previously glaciated parts of the Holarctic (including the brackish Baltic Sea), where the taxa are known as the "glacial relicts." The level of taxonomic separation among the Caspian, lacustrine, and marine forms varies from genus to genus, from no distinction within Limnocalanus macrurus (see Holmquist 1970) to suggested subgeneric and generic differences in Gammaracanthus (see Bousfield 1989). In Mysis, the morphologically and ecologically diversified Caspian species flock is paralleled by a lacustrine and brackish-water Mysis relicta group, which is morphologically rather uniform throughout the Holarctic (Holmquist 1959), but comprises four allozymically identified, largely allopatric sibling species, three of which occur in northern Europe (Vainola 1986; Vainola et al. 1994). Another four con-subgeneric (morphological) Mysis spp. live in northern marine waters (Holmquist 1959; Vainola 1992).
Various views have been presented on the relationships of the three ecological groups in the "relict" genera and on the timing of phylogenetic and zoogeographic events, particularly the colonization of inland waters (see Holmquist 1966; Segerstrale 1976). The two main lines of thought on the latter issue may be characterized as a Tertiary and a Pleistocene (Quaternary) hypothesis of origin. Underlying both concepts is the weak dispersal capacity of the animals. The freshwater species do not disperse actively upstream, nor passively with external vectors: their ranges are apparently determined by former direct water connections, from the sea and from and among proglacial (ice-dammed) lakes that emerged during the Pleistocene glaciations.
The Tertiary hypothesis assumes an invasion along an ancient direct seaway from the Arctic to the Ponto-Caspian basin. Sars (1927), who described much of the Caspian crustacean diversity, thought that the Arctic element was older than the prevalent autochthonous component of the fauna and had immigrated through a northern connection in a remote geological period. Holmquist (1959, 1966) claimed morphological evidence for a monophyly of the Caspian and lacustrine Mysis spp. From the wide freshwater distribution of M. relicta Loven, 1862, and from arguments on evolutionary rate, she proposed an early invasion of inland waters; a minimum age was obtained from the mid-Tertiary closure of the Turgai Strait (Obik Sea), which earlier linked the Arctic basin directly through the Aralo-Caspian region to the southern Tethys Sea. An ancient origin of the "relict" element has also been advocated by Bousfield (1989), and as regards the relationships between the marine and freshwater taxa, a pre-Pleistocene divergence is consistent with allozyme data (Vainola 1986; Vainola and Varvio 1989).
According to the Pleistocene hypothesis, the origin and dispersal of the nonmarine taxa were associated with proglacial lakes. It has been assumed that the boreal "relicts" were directly derived from marine ancestors in connection with one of the latest glaciations, as a glacier expanding from the north isolated estuarine populations to continental ice-dammed lakes (e.g. Hogbom 1917; Thienemann 1950; Segerstrale 1957). Although a recent relationship between the extant marine and freshwater taxa would be rejected, it is still obvious that the freshwater species did inhabit such ice-dammed refugial lakes during the glaciations, and have from there reinvaded their current northern lacustrine ranges (e.g., Kuderskii 1971; Dadswell 1974; Segerstrale 1982; Vainola et al. 1994). Moreover, it is evident that, during the glacial maxima, the drainage of the ice-dammed lakes shifted to the south toward the Caspian and Black Sea basins (e g., Grosswald 1980; Kvasov 1987), thus providing the freshwater populations a recent access to the Caspian Sea (see fig. 1). The hypothesis of such a late-glacial immigration over the current continental watershed--suggested by Hogbom (1917) and elaborated by, for example, Berg (1928), Pirozhnikov (1937), and Segerstrale (1957)-has survived as the prevailing concept on the origin of the Caspian Arctic element (Zenkevitch 1963; Kasymov 1987; Petryashov 1989). The Pleistocene flooding hypothesis could also be invoked to explain the evolution of the intra-Caspian Mysis species diversity: the recurring Middle and Late Pleistocene glacial cycles would have permitted recurring invasions from the northern lakes or seas, with intervening periods to allow evolution of reproductive isolation (cf. Hutchinson 1967).
In terms of phylogenetic relationships, the main hypotheses addressed here are the monophyly of the two nonmarine species groups (Caspian and freshwater) with respect to the marine "ancestors", and the monophyly of the Caspian flock itself (single or multiple colonizations). The evaluation of the temporal framework will be based on an assumption of a rough molecular clock, generalizing to mysids the rough correspondence of unit genetic distance with 5-20-my divergence time estimated in other animal groups (e.g., Thorpe 1983; Nei 1987; Grant 1987; Scheepmaker 1990). Even if it is applicable to the order of magnitude at best, the approach should retain some power for discriminating between the widely different time scales of the Tertiary and Pleistocene colonization hypotheses (for stronger reservations on rate calibrations, see Hillis and Moritz 1990; Bermingham and Lessios 1993).
Material and Methods
Three of the four Caspian Mysis species were collected from the southern basin of the sea, within 100 km southeast of Baku, in June 1991. Mysis caspia Sars, 1895, was obtained with an epibenthic beam trawl from about 100 m depth, and the planktonic species M. microphthalma Sars, 1895, and M. amblyops Sars, 1907, with vertical net hauls from 300 m at a 400-m site.
Information on samples of the four freshwater and five marine Mysis spp. used in the study is given elsewhere (Vainola 1992; Vainola et al. 1994). In the M. relicta species group, sp. I and sp. II are distributed in North European lakes and the Baltic Sea, sp. III is known from a single North Fennoscandian lake close to the Barents Sea coast, and sp. IV from lakes across continental northern North America. Mysis oculata Fabricius, 1780 and M. litoralis Banner, 1948 are Arctic-subarctic marine circumpolar species, both represented by amphiatlantic samples in the material. Mysis gaspensis Tattersall, 1954, and M. stenolepis Smith, 1873, are littoral species of the North American Atlantic coast. Mysis mixta Lilljoborg, 1852, is an Atlantic marine species; the material studied comprises Baltic, Barents, and White Sea samples. Excepting M. (Michteimysis) mixta and M. (Auricomysis) stenolepis, the species are referred to the nominate subgenus Mysis (see Vainola 1992).
The animals were frozen in liquid nitrogen and stored at -70[degrees]C until analysis. Horizontal starch-gel electrophoresis was performed as described in Vainola (1992). The Caspian species were directly cross-compared with material from all the previously studied marine and freshwater taxa. Seventeen enzyme loci were scored: Aco-2 (encoding aconitase), Ark1, Ark-2 (arginine kinase), Dia-2 (NADH diaphorase), Eno (enolase), Gapd (glycerol-3-phosphate dehydrogenase), Got1, Got-2 (glutamate-oxaloacetate transaminase), Gpi (glucosephosphate isomerase), Gpt (glutamate-pyruvate transaminase), Idh-1, Idh-2 (isocitrate dehydrogenase), Mdh-2 (malate dehydrogenase), Mpi (mannosephosphate isomerase), Pep-2 (dipeptidase), Pgm (phosphoglucomutase) and Tpi (triosephosphate isomerase). Of the additional loci in Vainola (1992), Pgk and Sdh were omitted here because of insufficient enzyme activity, and Dia-1 and Pgd because of excessive allelic diversity in the Caspian species, rendering allozyme identity assessments impractical. The reduction of loci had little effect on the overall pattern of interspecific relationships among the non-Caspian species of the subgenus Mysis as assessed in the earlier study (Vainola 1992).
Interspecific differences were summarized using Nei's genetic distance measures D = -logI (where I is genetic identity), and [D.sub.v] = (1 - I)/I, which is designed to be more linear with time when rate of change varies among loci (e.g., Nei 1987). Phenograms were constructed with the UPGMA (Sneath and Sokal 1973) and neighbor-joining (NJ) algorithms (Saitou and Nei 1987); the latter method, which does not imply equal branch lengths for sister groups, will more closely reflect the estimated distance matrix. Phylogenetic relationships were also examined in a qualitative character analysis with a locus as the character (e.g., Mooi 1989; Murphy 1993); M. (Michteimysis) mixta and M. (Auricomysis) stenolepis were used as initial outgroups for species in the subgenus Mysis (see Vainola 1992).
The three Caspian species are genetically very close to each other, but well separated from all marine and freshwater Mysis species. Estimates of genetic distance among M. caspia, M. microphthalma and M. amblyops (D [approximate] [D.sub.v] [approximate] 0.06) are an order of magnitude smaller than those to the extra-Caspian species in the same subgenus (D = 0.6-1.0; or [D.sub.v] = 0.81.9; table 1). The overall patterns of differentiation at 17 loci studied from the 12 species are illustrated in figure 2.
[TABULAR DATA 1 OMITTED]
No single allozyme locus is entirely diagnostic between any Caspian species pair (table 2). Yet, each species has private alleles at considerable frequencies; autapomorphs characteristic for a species are often found at loci that are monomorphic in the two others (Mdh-2 for M. caspia, Gpi and Gpt for M. microphthalma, Idh-2 and Pep-2 for M. amblyops). At Got-1 and Pgm, multiple alleles are shared by all three species, but one species differs from others in allele frequencies. Taken together, the studied characters do distinguish all species pairs with almost no ambiguity even at the individual level: with Hardy-Weinberg genotypic proportions and independent loci, the expected multilocus genotypic overlaps for pairwise species comparisons are 2-6 x [10.sup.-3] (no deviations from the HW genotypic structure were found within the species). The levels of intraspecific gene diversity are similar in the three species (H = 0.11 - 0.12), and not lower than those in other Mysis species (in four freshwater species H = 0.06-0.09, in five marine species H = 0.050.11, as calculated from the mean frequencies in each species at the 17 loci studied here, from data in Vainola ).
[TABULAR DATA 2 OMITTED]
The Caspian species group is distinguished from the extra-Caspian congeners by a set of uniquely shared alleles; putative Caspian synapomorphs are exclusive at Dia-2 and Gpt, and common at Mpi and Pgm (table 2). At most loci, however, the Caspian group shares alleles with some of the marine and freshwater con-subgeners while being diagnostically distinct from others. At the polymorphic Got-1, Pep-2, and Pgm, alternative Caspian alleles are present in different outgroup species. Allele sharing with the freshwater species is somewhat more common than with the marine congeners: 11-12 Caspian alleles are present in each of the M. relicta group species, 8-9 in marine species of subgenus Mysis, and 4-5 in the other subgenera (table 2). This is reflected in the distance estimates (table 1): distances from the Caspian to the freshwater species are, on average, lower than to the marine Mysis s. qtr. species (mean D = 0.7 vs. 0.9). However, as indicated by the phenograms (fig. 2), the pattern is not consistent enough to corroborate a monophyly of the composite nonmarine group (Caspian + freshwater). On the other hand, the other main hypothesis--the monophyly of the three Caspian species--is supported in the UPGMA phenogram at a 99.9% level when the distance data are bootstrapped over loci (Pamilo 1990) (for evaluation of other aspects of the branching pattern in the phenogram, see Vainola 1992).
No support for a Caspian + freshwater clade within subgenus Mysis can either be claimed from a direct assessment of allele distributions in an outgroup comparison to the other subgenera. The observed patterns of broad allele sharing in Mysis species (table 2; Vainola 1992) generally imply retention of polymorphisms through multiple branching events and/or homoplastic allele origins at several loci with any cladogram (e.g., at Got-1, Got-2, Mpi, Pgm, Tpi, which are often polymorphic in present populations also). The phylogenetic information provided by a number of less polymorphic loci is also often contradictory. In the earlier analysis of relationships among the non-Caspian conventionally recognized taxa, the best cladogram was based on only three compatible informative characters (Ark-2, Eno, Pep-2); all these supported an oculata-litoralis clade but left the relative positions of the relicta group and gaspensis unresolved (Vainola 1992). With the extended set of taxa, two of these loci (Eno, Pep-2) do involve alleles uniquely shared by the Caspian and freshwater groups; nevertheless, the outgroup approach suggests a branching pattern similar to that of figure 2b, where gaspensis is situated basal to the others, and the Caspian clade is linked as a sister group to the terminal oculata-litoralis clade (by Ark-[2.sup.104]), rather than to the M. relicta group. At any rate, even this hypothesis implies homoplasy at a number of potentially informative loci. The extension of the data also affects inferences on character state polarity that earlier seemed informative with regard to the relationships within the M. relicta group (Vainola et al. 1994). Assuming a monophyly of the M. relicta group (which in itself has no direct molecular cladistic support), the evidence for a terminal species I+IV clade is weakened as two of the three putative synapomorphic alleles, Got-[1.sup.100] and Pep-[2.sup.100], are encountered in the Caspian group.
Molecular and Morphological Divergence in an Endemic Species Flock
The data confirm the monophyly of the Caspian Mysis species group (as represented by the three species studied), and suggest remarkably close evolutionary relationships among the endemic taxa. Yet the results also clearly demonstrate the genetic distinctness and reproductive isolation among these largely sympatric species.
The molecular uniformity among the morphologically well separated Caspian Mysis spp. is striking when compared with the contrasting patterns of molecular versus morphological divergence in species from the two other zoogeographical zones. An indirect but illustrative account of the morphological differences within each of the Caspian, lacustrine, and marine Mysis species groups is given by their taxonomic records. The four Caspian species were distinguished at the turn of the century (Sars 1907), whereas the morphologically recognized marine species were only separated in the 1950s (see Holmquist 1959), and the sibling species of the freshwater M. relicta group still await a morphological description. In addition to the more conventional taxonomic characters differences generally related to their ecology and life style. Mysis caspia is a customarily robust and large-eyed nektobenthic Mysis species, whereas the planktonic M. microphthalma and M. amblyops have greatly reduced eyes, the former being a very slender species with long appendages, and the latter an exceptionally small one (ca. 12 mm length in contrast to 25-30 mm in other Mysis spp.) with proportionally reduced abdominal segments (Sars 1907).
With the assumed time correlation, the observed differences (D = 0.06) indicate a recent speciation and rapid morphological differentiation for the sympatric Caspian species group, most likely in the Middle or Late Pleistocene, possibly within the last few glaciation cycles. By contrast, the boreal lacustrine species, largely allopatric and with greatly fragmented population structures, exhibit an apparent morphological stasis over a considerably longer period (D [approximate] 0.4). In the earlier comparison of marine and freshwater taxa, the morphological stability in lakes seemed notable even in view of the (rather slight) differentiation among the somewhat older marine species (D [approximate] 0.8). From the present data, the variation of evolutionary rates appears even greater.
The recency of the Caspian divergence provides a parallel to that of the species-rich haplochromine cichlid flock of Lake Victoria, where allozyme distances have been estimated at D = 0-0.04 (Sage et al. 1984), and a young age, cat 200,000 yr, has also been suggested from mtDNA data (Meyer et al. 1990). Similarly close allozyme relationships are known from other, moderately diversified intralacustrine fish species flocks (Humphries 1984; Kornfield and Carpenter 1984). In a broader context, though, the intraflock genetic differences both in fishes and Mysis represent a level typical of conspecific allopatric populations rather than of distinct taxa in the animal kingdom (e.g., Thorpe 1983). In the boreal congeners, similar distances are found between allopatric populations within individual M. relicta group species. Although the latter differences probably result from isolation in disjunct glacial refugia and may reflect divergence times on the order of 105 years, they obviously do not represent biological speciation (Vainola et al. 1994). In marine Mysis spp., estimates of transoceanic intraspecific differentiation are somewhat lower, D < 0.03 (Vainola 1992).
Much of the diversity in ancient lakes is currently attributed to intralacustrine speciation, though differentiation in populations earlier isolated by absolute geological barriers may also have had a role (Brooks 1950; Fryer and Iles 1972; Mayr 1984). Allopatric intralacustrine divergence is thought to be facilitated by geographical isolation (distance) within large lakes and by more regional habitat discontinuities. Recent discussions also emphasize the importance of sexual selection, local trophic specialization, and resource allocation through feeding and breeding allopatry or allochrony in enhancing the spatial segregation of populations, which is essential in protecting the evolution and maintenance of the distinct ecological and morphological identities of the emerging taxa (e.g., Mayr 1984; Smith and Todd 1984; Fryer 1991). Intralacustrine divergence is noted to be most typical of organisms with morphological and life history traits conferring weak dispersal capacity and small local population sizes (Cohen and Johnston 1987). Accordingly, remarkably little diversification is seen in the pelagic fauna of ancient freshwater lakes, in contrast to the diversity in their littoral zones (e.g., Brooks 1950).
Against this background, the Caspian Mysis flock appears exceptional. At least two species (M. microphthalma and M. amblyops) are almost entirely planktonic and apparently sympatric in the deep-water zone; M. caspia also spends much of its life in the plankton, but mainly reproduces in the near-bottom waters (Bondarenko 1991). Although the species may currently effectively divide the pelagic resources (e.g., with regard to prey size and depth zone), and selection for such partitioning could have driven the morphological specialization, it is hard to envisage conditions that would have promoted the accumulation of differences by impeding contacts of differentiating lineages. Pelagic mysids should not be prone to any spatial genetic substructuring within a continuous basin; this is seen, for example, in the large-scale population homogeneity in Mysis spp. within the Baltic Sea (Vainola 1992). The Caspian mysids show no reduction in heterozygosity or allele number to suggest a role for a major bottleneck in speciation; this is also the case with other species flocks (e.g., Sage et al. 1984). Interestingly, even though the diversity of Caspian Mysis makes a contrast to the poverty of larger planktonic crustaceans in ancient freshwater lakes, the diversity is not unique within the Caspian itself: a flock of endemic polyphemid cladocerans also inhabits the open Caspian waters (e.g., Birstein et al. 1968).
Temporary subdivision into separate water bodies by lowered water levels is one of the more classical schemes invoked to enable allopatric speciation of intralacustrine congeners (e.g., Brooks 1950; Fryer and Iles 1972; Sturmbauer and Meyer 1992). Although the level of the Caspian has fluctuated widely, with consequent effects on lake morphology and chemistry, the regressions during the Middle and Late Pleistocene were probably not great enough to completely isolate the distinctive southern and central subbasins of the lake, now separated by a sill at -- 200 m (Fedorov 1978). Rather the transgressions caused by increased glacial runoff, raising the lake to up to 75 m over its present level and expanding its area far to the northern lowlands (fig. 1), might have created new environments supporting the diversification of coldwater Mysis spp. (cf. Kvasov 1987).
On the one hand, there are clear parallels in the molecular data from species flocks elsewhere and from the Caspian Mysis. On the other hand, there are parallels in the intrabasin diversity of Arctic Caspian Mysis and in genera representing the main autochthonous temperate element of the Caspian fauna. Yet, considering the fundamentally different zoogeographical and ecological attributes of the two Caspian elements, a direct generalization from the recent divergence of Mysis to the age of autochthonous flocks might not be warranted. Much of the Caspian autochthonous diversity may still be relatively recent. Although some forms congeneric or even conspecific with the present ones are known from the Late Miocene, the present type of fauna finally replaced an earlier one of a more saline character only in the Early Pleistocene ( <1.6 mya), possibly by immigration from westerly parts of the Ponto-Caspian system (e.g., Kasymov 1987). Direct fossil evidence of recent evolution of Caspian bivalves is found, for example, in the succession of Didacna species in the Middle and Late Pleistocene (Fedorov 1978).
Origin of the Caspian Arctic Element
The close intra-Caspian relationships in Mysis are contrasted by a lack of any comparable affinities to the extra-Caspian congeners. With the assumed time correlation, the data give no support for a descent of the Caspian Arctic species from any of the studied lacustrine or marine lineages in the Late or Middle Pleistocene, that is, in the period of major continental glaciations associated with a southward drainage of northern proglacial waters (10,000-900,000 yr ago). Particularly, there is no indication of a derivation of the sympatric Mysis flock from an extra-Caspian source by serial descent and reintroductions during successive glaciations or by multiple colonizations by different extra-Caspian ancestors--another traditional hypothesis for an allopatric genesis of a species flock (Brooks 1950).
The Pleistocene colonization hypothesis has been presented in several versions regarding the immigration route and the phylogenetic relationships of the three zoogeographical groups. The original and most popular variant suggested a recent immigration from ice-dammed lakes in Eastern Europe, through the Volga river (e.g., Hogbom 1917; Berg 1928; Zenkevitch 1963). The model initially also assumed a Late Pleistocene descent from marine congeners, not supported by allozyme data (Vainola 1986, 1992). Alternatively, the lacustrine taxa were thought to have evolved in northern estuaries in the Early or Middle Pleistocene, before the colonization of continental waters (Dadswell 1974; Golikov and Scarlato 1989). Anyway, disregarding the marine relationship, the scheme of a northwesterly continental origin of the Caspian Arctic element would appear geographically sound: paleohydrographical reconstructions (e.g., Grosswald 1981); Kvasov 1987) combined with distributional evidence suggest that during the latest glaciation, two species of the M. relicta group were present in northeastern European refugial lakes that discharged to the Caspian Sea through the Volga River (Vainola et al. 1994; see fig. 1). As the potential ancestors are thus identified, but do not show close relationship to the Caspian taxa, this model may be confidently abandoned.
On the other hand, similar proglacial impoundments and drainage diversions toward the Caspian basin also took place in the West Siberian lowlands east of the Urals, although not as late as the last glacial maximum, cat 20,000 yr ago (see Faustova 1984; Astakhov 1987). This route for a Pleistocene colonization was advocated by Pirozhnikov (1937) and Segerstrale (1957, 1976). The identity of the M. relicta group species from the Siberian coastal sector has not been electrophoretically assessed. In addition, considering the evidence of cryptic systematic diversity and the limited phylogenetic resolution, a derivation from an unknown extraCaspian Mysis lineage earlier in the Pleistocene perhaps cannot be as strictly excluded. At any rate, the studied material comprises all the morphologically closest known relatives of the Caspian flock (Holmquist 1959).
Although the prevailing view of recent, Pleistocene relationships of the vicarious taxa could be rejected, the data do not immediately suggest a plausible alternative. The rival hypothesis involving immigration to the Caspian through a direct Tertiary northern marine connection (Holmquist 1966) appears equally unlikely. The Turgai Strait was probably closed already in the late Eocene-Oligocene, 40-30 mya (Marinovich et al. 1990), but the genetic data would more reasonably fit divergences some 3-15 mya. The hypothesis seems even more improbable in view of the climate and environment at that time, with open connections to the Tethys in the south, before the evolution of boreal elements in the north (Golikov and Scarlato 1989) and much before the emergence of the autochthonous Caspian brackish-water fauna. Assuming that the cold-water habit shared by the Caspian Arctic element and the northern congeners is monophyletic, it would thus be more reasonable to adjust the concepts of divergence rate to comply with an early glacial event than with the mid-Tertiary seaway. However, a solution may as well require a reconsideration of the Pliocene history of northerly connections (cf. Steininger and Rogl 1984). An alternative western marine route through the Mediterranean has been deemed improbable as no traces of other truly boreal-Arctic taxa are seen in the fossil records of the PontoCaspian region (Hogbom 1917; Zubakov 1990). Although there is evidence that the climatic and eustatic fluctuations during the Plio-Pleistocene repeatedly created connections and faunal exchanges among the Mediterranean, Black Sea, and Caspian basins, the marine incursions from the west coincided with the warm periods with high water levels and temperate faunas (Zubakov 1988, 1990).
When considering the colonization and subsequent survival of the Arctic element in the Caspian basin, it is notable that all the lacustrine "relict" taxa with marine affinities have vicarious forms in the Caspian. Moreover, the Caspian element includes one northern marine crustacean genus (Onisimus) not present in boreal lakes. A credible zoogeographical history of this element can hardly involve anything like chance events, including catastrophic environmental variations ensuing high extinction risk. Therefore, a very early immigration would seem unlikely. And if the Mysis lineage had long been present in the basin, with little risk of extinction, why would the radiation of the (surviving) flock have occurred only so recently? Assuming a phylogeny reflected in figure 2a, and a recent final radiation within the basin, the branch leading to the Caspian group would represent a long period that escapes zoogeographical inferences. Similar studies of the other members of the Arctic element may help in bracketing the colonization event; closer extra-Caspian affinities in these taxa would support the view that the distant relationship in Mysis actually reflects a descent from an unrecognized extra-Caspian lineage.
The comparative data on the three zoogeographical and ecological groups of Mysis species provide another example of evolutionary dynamics in which morphological and molecular differentiation and the speciation process appear to be related in no consistent pattern. In the pelagic species flock of the Caspian Sea, both speciation and morphological divergence have been rapid. They have probably been related to ecological specialization, but from the geographical settings, it is hard to envisage the role of allopatry in protecting the process.
On the other hand, in the boreal zone, the fragmented lacustrine environment and the repeated distributional changes associated with glacial cycles would seem to have provided ample opportunities for allopatric differentiation. In this framework, however, a common M. relicta morphology has been maintained for an extensive period. Biological speciation has taken place in some instances and may be associated with some ecological differentiation, such as in salinity tolerance, but even this divergence has probably arisen on a broader time scale than that in the Caspian Sea. More recent refugial distinctions, reflected in molecular differences of the same order as the interspecific Caspian distinctions, do not seem to correspond to biological species within the M. relicta group (Vainola et al. 1994).
In the marine environment, relative genetic continuity has been maintained over larger distances (i.e., transoceanic) than in the boreal lacustrine complex (Vainola 1992). However, the earlier suggestion of the importance of the fragmented lacustrine environment in promoting speciation retains little generality with the present indication of intralacustrine divergence in the Caspian complex.
The prevailing view of a recent Late Pleistocene colonization of the Caspian basin by the Arctic element through Eastern European continental proglacial lakes that probably served as refugia for the boreal lacustrine "glacial relict" crustaceans was rejected. In general, no strong evidence for a common descent of the vicarious Caspian and boreal lacustrine crustacean assemblages was obtained. The divergence of all the three distributional groups seems to have been ancient, although not as old (mid-Tertiary) as suggested by some earlier authors (Sars 1927; Holmquist 1966).
I thank A. G. Kasymov and V. M. Gasanov of the Zoological Institute of the Azerbaijan Academy of Sciences for arranging sampling on r/v Elm, C. Olsson for help with electrophoresis, and the referees for valuable comments. The study was supported by the Academy of Finland.
[Figures 1&2 ILLUSTRATION OMITTED]
Arkhipov, S. A., V. G. Bespaly, M. A. Faustova, O. Yu. Glushkova, L. A. Isaeva, and A. A. Velichko. 1986. Ice-sheet reconstructions. Quaternary Science Reviews 5475-483.
Astakhov, V. 1987. Origin of West Siberian lakes. Pp. 144-155 in A. Raukas and L. Saarse, eds. Palaeohydrology of the temperate zone, Vol. 1. Rivers and lakes. Valgus, Tallinn, Estonia.
Berg, L. S. 1928. O proiskhozhdenii severnykh elementov v faune Kaspiya. Doklady Akademii Nauk SSSR A 7:107-112.
Bermingham, E. and H. A. Lessios. 1993. Rate variation of protein and mitochondrial DNA evolution as revealed by sea urchins separated by Istmus of Panama. Proceedings of the National Academy of Sciences, USA 90:2734-2738.
Birstein, Ya. A., L. G. Vinogradov, N. N Kondrakov, M. S. Kun, T. V. Astakhova, and N. N. Romanova, eds. 1968. Atlas Bespozvonochnykh Kaspiiskogo Morya. Izdatelstvo Pishchevaya Promyshlennost, Moscow.
Bondarenko, M. V. 1991. Reproduktivnye tsikly planktonnykh mizid Kaspiiskogo morya. Pp. 147-157 in V. I. Kuzmicheva, ed. Rybokhozyaistvennve Issledovaniya Planktona. II. Kaspiiskoe More: Sbornik Plauchnykh Trudov. VNIRO, Moscow.
Bousfield, E. L. 1989. Revised morphological relationships within the amphipod genera Pontoporeia and Gammaracanthus and the "glacial relict" significance of their postglacial distributions. Canadian Journal of Fisheries and Aquatic Sciences 46:1714-1775.
Brooks, J. L. 1950. Speciation in ancient lakes. Quarterly Review of Biology 25:30-60, 131-176.
Cohen, A. S., and M. R. Johnston. 1987. Speciation in brooding and poorly dispersing lacustrine organisms. Palaios 2:426-435.
Dadswell, M. J. 1974. Distribution, ecology, and postglacial dispersal of certain crustaceans and fishes in eastern North America. National Museum of Natural Sciences. Publications in Zoology 11:1-110.
Echelle, A. A., and I. Kornfield, eds. 1984. Evolution of fish species flocks. University of Maine at Orono Press, Orono.
Faustova, M. A. 1984. late Pleistocene glaciation of European USSR. Pp. 3-12 in A. A. Velichko, ed. Late Quaternary environments of the Soviet Union. Longman, London.
Fedorov, P.V. 1978. Pleistotsen Ponto-Kaspiya. Trudy Geologicheskogo Instituta 310. Nauka, Moscow.
Fryer, G. 1991. Comparative aspects of adaptive radiation and speciation in Lake Baikal and the great rift lakes of Africa. Hydrobiologia 211:137-146.
Fryer, G., and T. D. Iles. 1972. The cichlid fishes of the great lakes of Africa. Oliver and Boyd, Edinburgh.
Golikov, A. N., and O. A. Scarlato. 1989. Evolution of Arctic ecosystems during the Neogene period. Pp. 357-279 in Y. Herman, ed. The Arctic seas. Van Nostrand Reinhold, New York.
Grant, W. S. 1987. Genetic divergence between congeneric Atlantic and Pacific Ocean fishes. Pp. 225-246 in N. Ryman and F. Utter, eds. Population genetics and fishery management. Washington University Press, Seattle.
Grosswald, M. G. 1980. Late Weichselian ice sheet of northern Eurasia. Quaternary Research 13:1-32.
Hillis, D. M., and C. Moritz. 1990. An overview of applications of molecular systematics. Pp. 502-515 in D.M. Hillis and C. Moritz, eds. Molecular systematics. Sinauer, Sunderland, Mass.
Hogbom, A. G. 1917. Uber die arktischen Elemente in der aralokasprschen Fauna, ein tiergeoraphisches Problem. Bulletin of the Geological Institutions of the University of Uppsala 14:24-260.
Holmquist, C. 1959. Problems on marine-glacial relicts on account of investigations on the genus Mysis. Berlingska Boktryckeriet, Lund, Sweden.
--. 1966. Die sogenannten marin-glazialen Relikte nach neueren Gesichtspunkten. Archiv fur Hydrobiologie 62:285-326.
--. 1970. The genus Limnocalanus. Zeitschrift fur Zoologische Systematik und Evolutionsforschung 8:273-296.
Humphries, J. M. 1984. Genetics of speciation in pupfishes from Laguna Chichancanab, Mexico. Pp. 129-139 in A. A. Echelle and I. Kornfield, eds. Evolution of fish species flocks. University of Maine at Orono Press, Orono.
Hutchinson, G. E. 1967. A treatise on limnology, Vol. 2. Introduction to lake biology and the limnoplankton. Wiley, New York.
Kasymov, A. G. 1987. Zhivotnyi Mir Kaspiiskogo Morya. Isdatelstvo Elm, Baku, Azerbaijan.
Kornfield, I., and K. E. Carpenter. 1984. Cyprinids of Lake Lanao, Philippines: Taxonomic validity, evolutionary rates and speciation scenarios. Pp. 69-84 in A. A. Echelle and I. Kornfield, eds. Evolution of fish species flocks. University of Maine at Orono Press, Orono.
Kozhov, M. 1963. Lake Baikal and its life. Junk, The Hague.
Kuderskii, L. A. 1971. O proiskhozhdenii reliktovoi fauny v ozerakh severo-zapada evropeiskoi chasti SSSR. Izvestiya Gosudarstvennogo Nauchno-Issledovatel skogo Instituta Ozernogo i Rechnogo Rybnogo Khozyaistva 76:113-123.
Kvasov, D. 1987. Development of northern and central Eurasian lakes in the Late Quaternary. Pp. 156-168 in A. Raukas and L. Saarse, eds. Palaeohydrology of the temperate zone, Vol. 1. Rivers and lakes. Valgus, Tallinn, Estonia.
Marinovich, L., Jr., E. M. Brouwers, D. M. Hopkins, and M. C. McKenna. 1990. Late Mesozaic and Cenozoic paleogeographic and paleoclimatic history of the Arctic Ocean Basin, based on shallow-water marine faunas and terrestrial vertebrates. Pp.403-426 in A. Grantz, L. Johnson, and A. Sweeney, eds. The Arctic Ocean region. Geology of North America, Vol. L. The Geological Society of America, Boulder, Colo.
Mayr, E. 1984. Evolution of fish species flocks: a commentary. Pp. 185-202 in A. A. Echelle and 1. Kornfield, eds. Evolution of fish species flocks. University of Maine at Orono Press, Orono.
Meyer, A., T. D. Kocher, P. Basasibwaki, and A. C. Wilson. 1990. Monophyletic origin of Lake Victoria cichlid fishes suggested by mitochondrial DNA sequences. Nature 347:550-553.
Mooi, R. 1989. The outgroup criterion revisited via naked zones and alleles. Systematic Zoology 38:283-290.
Mordukhai-Boltovskoi, P.D. 1964. Caspian fauna beyond the Caspian Sea. Internationale Revue der Gesamten Hydrobiologie 49: 139-176.
Murphy, R. W. 1993. The phylogenetic analysis of allozyme data: invalidity of coding alleles by presence/absence and recommended procedures. Biochemical Systematics and Ecology 21: 25-38.
Nei, M. 1987. Molecular evolutionary genetics. Columbia University Press, New York.
Nishida, M. 1991. Lake Tanganyika as an evolutionary reservoir of old lineages of East African cichlid fishes: inferences from allozyme data. Experientia 47:974-979.
Pamilo, P. 1990. Statistical tests of phenograms based upon genetic distances. Evolution 44:689-697.
Petryashov, V. V. 1989. Arctic Ocean mysids (Crustacea: Mysidacea): evolution, composition and distribution. Pp. 373-396 in Y. Herman, eds., The Arctic seas. Van Nostrand Reinhold, New York.
Pirozhnikov, P. L. 1937. A contribution to the study of the origin of the northern elements in the fauna of the Caspian Sea. Doklady Akademii Nauk SSSR 15:521-524.
Ribbink, A. J. 1984. Is the species flock concept tenable? Pp. 2125 in A.A. Echelle and I. Kornfield, eds. Evolution of fish species flocks. University of Maine at Orono Press, Orono.
Sage, R.D., P.V. Loiselle, P. Basasibwaki, and A. C. Wilson. 1984. Molecular and morphological change among cichlid fishes of Lake Victoria. Pp. 185-202 in A. A. Echelle and I. Kornfield, eds. Evolution of fish species flocks. University of Maine at Orono Press, Orono.
Saitou. N., and M. Nei. 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4:406-425.
Sars, G. O. 1907. Mysidae. Pp. 242-313 in Trudy Kaspiiskoi Ekspeditsii 1904 Goda. Tom I. Byuro po Promyslovoi Zoologii i Rybovodstvu. St. Petersburg, Russia.
--. 1927. Notes on the crustacean fauna of the Caspian Sea. Pp. 315-329 in Sbornik v Chest Prof. N.M. Knipovicha. Moscow.
Scheepmaker, M. 1990. Genetic differentiation, origin and dispersal of Gammarus gauthieri from the Iberian peninsula and North Africa (Crustacea, Amphipoda). Bijdragen tot de Dierkunde 60:31-49.
Segerstrale, S. G. 1957. On immigration of the glacial relicts of Northern Europe, with remarks on their prehistory. Commentationes Biologicae Societas Scientiarum Fennica 16(16):1-117.
--. 1976. Proglacial lakes end the dispersal of glacial relicts. Commentationes Biologicae Societas Scientiarum Fennica 83: 1-15.
--. 1982. The immigration of glacial relicts into Northern Europe in the light of recent geological research. Fennia 160: 303-312.
Smith, G. R., and T. N. Todd. 1984. Evolution of species flocks of fishes in north temperate lakes. Pp. 45-68 in A. A. Echelle and I. Kornfield, eds. Evolution of fish species flocks. University of Maine at Orono Press, Orono.
Sneath, P. H. A., and R. R. Sokal. 1973. Numerical taxonomy. W. H. Freeman, San Francisco.
Steininger, F. F., and E Rogl. 1984. Paleogeography and palinspastic reconstruction of the Neogene of the Mediterranean and the Paratethys. Pp 659-674 in J. E. Dixon and A. H. E Robertson, eds. The geological evolution of the eastern Mediterranean. Blackwell Scientific, Oxford.
Sturmbauer, C., and A. Meyer. 1992. Genetic divergence, speciation and morphological stasis in a lineage of African cichlid fishes. Nature 358:578-581.
Thienemann, A. 1950. Verbreitungsgeschichte der Susswassertierwelt Europas. Die Binnengewasser 28:1-809.
Thorpe, J. P. 1983. Enzyme variation, genetic distance and evolutionary divergence in relation to levels of taxonomic separation. Pp. 131-152 in G.S. Oxford and D. Rollinson, eds. Protein polymorphism: adaptive and taxonomic significance. Academic Press, London.
Vainola, R. 1986. Sibling species and phylogenetic relationships of Mysis relicta (Crustacea: Mysidacea). Annales Zoologici Fennici 23:207-221.
--. 1992. Evolutionary genetics of marine Mysis spp. (Crustacea: Mysidacea). Marine Biology 114: 539-550.
Vainola, R., and S.-L. Varvio. 1989. Molecular divergence and evolutionary relationships in Pontoporeia (Crustacea: Amphipoda). Canadian Journal of Fisheries and Aquatic Sciences 46: 1705-1713.
Vainola, R., B.J. Riddoch, R.D. Ward, and R.I. Jones. 1994. Genetic zoogeography of the Mysis relicta species group (Crustacea: Mysidacea) in northern Europe and North America. Canadian Journal of Fisheries and Aquatic Sciences 51:1490-1505.
Zenkevitch, L. 1963. Biology of the seas of the USSR. Allen and Unwin, London.
Zubakov, V. A. 1988. Climatostratigraphic scheme of the Black Sea Pleistocene and its correlation with the oxygen-isctope scale and glacial events. Quaternary Research 29:1-24.
--. 1990. The climatostratigraphy of the Mediterranean Pliocene and terminal Miocene. International Geology Review 32: 878-889.
Risto Vainola, Present address: Zoological Museum, PO. Box 17, FIN-00014, University of Helsinki, Finland. E-mail: email@example.com
|Printer friendly Cite/link Email Feedback|
|Date:||Dec 1, 1995|
|Previous Article:||Retrodisplacement of the oral and anal openings in dendrasterid sand dollars.|
|Next Article:||Natural selection on quantitative traits in the Bombina hybrid zone.|