Antarctic Science 12 (3): 276-287 (2000) 0 British Antarctic Survey Printed in the United Kingdom

Antarctic notothenioid fishes as subjects for research in evolutionary biology

JOSEPH T. EASTMAN Department of Biomedical Sciences, Ohio University, Athens, OH 45701-2979, USA

eastman@ohiou edu

Abstract: Antarctica is a continental island and the waters of its shelfand upper slope are an insular evolutionary site. The shelf waters resemble a closed basin in the Southern Ocean, separated from other continents by distance, current patterns and subzero temperatures. The benthc fish fauna of the shelf and upper slope of the Antarctic Region includes 213 species with higher taxonomic diversity confined to 18 families. Ninety-six notothenioids, 67 liparids and 23 zoarcids comprise 45%, 32% and 11% of the fauna, a combined total of 88%. In high latitude (71-78”s) shelf areas notothenioids dominate abundance and biomass at levels of 90-95%. Notothenioids are also morphologically and ecologically diverse. Although they lack a swim bladder, the hallmark of the notothenioid radiation has been repeated diverslfication into water column habitats. There are pelagic, semipelagic, cryopelagc and epibenthic species. Notothenioids exhbit the Qsproportionate speciosity and high endemism characteristic offish species flock. Antifreeze glycopeptides originating from a transformed trypsinogen gene are a key innovation. It is not known when the modem Antarctic shelffauna assumed its current taxonomic composition. A late Eocene fossil fauna was taxonoinically diverse and cosmopolitan. There was a subsequent faunal replacement with little carryover of families into the modem fauna. Basal notothenioid clades probably diverged in Gondwanan shelf locations during the early Tertiary. Dates inferred from molecular sequences suggest that phyletically derived Antarctic clades arose 15-5 n1.y.a.

Received 5 October 1999, accepted 21 February 2000

Key words: adaptive radiation, Antarctic, evolution, fish fauna, Notothenioidei, species flock

Introduction

Evolution is the major unlfying principle of biology and evidence of the evolutionary process pervades all levels of biological organization from molecules to ecosystems. Although the influence of evolution extends to the fauna of the most isolated continent, Antarctica is not usually included among notable evolutionary sites such as the Galapagos, Hawaii, Australia, Madagascar, the East African Great Lakes and Lake Baikal. The Scientific Committee on Antarctic Research (SCAR) Workshop on “Evolutionary Biology of Antarctic Organisms” has highlighted the potential of the Antarctic fauna for evolutionary studies. Topics dealing with fishes were well represented among the papers given at the workshop. Perhaps more than any other group of organisms, the radation of notothenioids, a suborder of perciforms, has served as a subject for research in Antarctic evolutionary biology. After introductory comments on Antarctica as an evolutionary site, I will provide an overview of the fish fauna with a summary of the taxonomic composition, a basic prerequisite for evolutionary studies. Sections will follow on notothenioid biodiversity, the adaptive radiation of notothenioids and the Cenozoic fossil fish faunas.

Antarctica as an evolutionary site

Studies of evolutionary radiations on islands are naturally

focussed on terrestrial faunas that evolved in isolation. Antarctica is a continental-sized island, twice the size of Australia, with the dominant fauna inhabiting the waters of the adjacent continental shelf rather than the landmass. The insular characteristics of the shelfinclude geographic isolation by distance, oceanographic isolation by currents and thermal isolation by subzero temperature. Although the Antarctic is not mentioned in the recent book Evalution an Islands, Grant (1998, p. vi) notes that the attention ofevolutionarybiologists is drawn to islands, as well as isolated lakes, by a number of common features ofisland faunasincludingtheir “strangeness”, replicated evolution and rapiQty of diverslfication. Strangeness or uniqueness certainly characterizes notothenioid fishes living at subzero temperatures with antifreeze and without haemoglobin; fishes that, although lacking a swim bladder, have repeatedly diversified into water column niches; fishes that are dominant at abundance and biomass levels of 90-95% and fishes that have a history estimated in millions rather in than tens of millions of years.

Ichthyologists have had access to the Antarctic shelffauiia for a century and therefore, unlike the deep-sea and large tropical river systems, this fauna is well characterized taxonomically (Gon & Heemstra 1990, Miller 1993, Andriashev 1986, Andriashev & Stein 1998) and also from otherbiological perspectives (Andriashev 1965, DeWitt 197 1,

276

277 ANTARCTIC FISH FAUNA

Balushkin 1984, Kock 1992, Eastman 1993, di Prisco et al. 1991, 1998). As thebest example of an adaptive radiationof marine fishes, the notothenioid diverslfication hasbeen studied at the organismal, organ system and molecular levels of organization (Eastman 1993, Bargelloni etal. 1994, Clarke & Johnston 1996, Klingenberg& Ekau 1996, Ritchie etal. 1996, Chen et al. 1997a, 1997b, Cheng 1998a, di Prisco 1998, Eastman & Clarke 1998, Ozouf-Costaz et al. 1997, Pisano et al. 1998). However, notothenioids arejust beginning to be appreciated outside the Antarctic scientific community and to serve as exemplars of adaptive radiation (Smith 1996, p. 71, Givnish 1997, p. 21,30-31, Helfmanetal. 1997, p. 307-309, Johns & Avise 1998).

Perhaps the reason the Antarctic shelf waters and its fish fauna are under-appreciated in an evolutionary context is that much research to date, with the encouragement of funding agencies, has emphasized aspects of extreme biology rather than w i n g principles of evolutionary biology. This approach has yielded valuable information about molecular and organ system function at low temperature, but has not focussed attention on aspects of organismal biology that the Antarctic fauna shares with faunas in other isolated habitats such as rift lakes, postglacial lakes and islands. Even among biologists there is little notoriety associated with a fauna that is remote and difficult to access through ice-covered shelf waters. The evidence for evolution is there, however, and research on the fish fauna will provide insight into evolutionary processes in marine habitats as well as raising the visibility of Antarctic evolutionary biology.

The modern environment and fish fauna

The Southern Ocean is the expanse of cold sea-water south of the mean position of the Antarctic Polar Front (45-60°S), including the Sub-Antarctic islands of the Indian Ocean sector (Gon & Heemstra 1990, p. 70). This is the Antarctic Zoogeographic Region. The Southern Ocean encompasses about one-tenth of the world ocean and the Antarctic continental shelf, where most of the fishes are found, comprises 2 x 106km2 (Pakhomov 1997, p. 378), about 7-8%ofthe 27-29 x lo6 km2 of total shelf area in the world (Kennett 1982). Unllke other large marine ecosystems, the shelf and upper slope of the Antarctic Region resemble a closed basin, isolated from other shelf areas by distance, current patterns and subzero temperatures. Thus the ranges of many fish species, while large and frequently encompassing the entire continental shelf, are as circumscribed as those of lacustrine endemics elsewhere in the world. The fish fauna is also hghly endemic, with 88% of the species confined to waters south of the AntarcticPolarFront (Andriashev 1987). Ifjust notothenioids are considered, species endemism rises to 97% (Andriashev 1987). Comparative figures for endemism of near-shore ichthyofaunas in other isolated marine habitats include 9-16% for the Galapagos, 23% for Easter Island and 25% for Hawaii (Grove & Lavenberg 1997, McCosker 1998).

Number of species

Fishes of the Southern Ocean (Gon & Heemstra 1990), the definitive work on Antarctic fishes, inventoried the fauna of the Antarctic Region at 272 species representing 49 families. In the intervening decade this number has risen to 3 13 species from 50 families. This is 1.2-1.3% of the world's fish fauna of 24 618 (Nelson 1994) to 25 000 (Eschmeyer 1998) valid species. While some of the 41 additions are new species, others are simply new records for the Southern Ocean. For example, a complete treatment of the liparid genus Careproctus (Andnashev & Stein 1998) was not available for inclusion in Fishes of the Southern Ocean, With the publication of this monograph on Careproctus, the most noteworthy change in the past decade has been an increase in the count of Southern Ocean liparids from 31 to 67 species. This family has surpassed the Nototheniidae as the most speciose family of Antarctic fishes.

Benthic fish are the major component of the fauna on the shelfand upper slope of the Antarctic Region. There are 213 benthic species and hlgher taxonomic diversity is confined to 18 families (Table I). In the high latitude (71-78"s) embayments of the Weddell and Ross seas, the fauna consists of about 80 species (Hubold 1992, Eastman & Hubold 1999). Ths includes the endemic shelfcomponent and a cosmopolitan non-notothenioid mesopelagic component (not included in Table 1) - the latter consisting of less than 10 species in the Weddell and Ross Seas (Hubold & Ekau 1987, Eastman & Hubold 1999). Representatives of this cosmopolitan group includebathylagids, paralepidds, myctophids and macrourids.

Table I. Families of benthic fishes from the Antarctic Region.

Taxona

Myxinidae (hagfkhes) 1 0.5 Petromyzontidae (lampreys) 1 0.5

Carapidae (pearlfishes) 1 0.5

No of speciesb % fauna

Rajidae (skates) 8 3.7

Moridae (deepsea cods) 4 1.8 Muraenolepididae (eel cods) 4 1.8 Gadidae (cods) 1 0.5 Congiopodidae (horsefishes) 1 0.5 Bathylutichthyidae' 1 0.5 Liparidae (snailfishes) 67 31.5 Zoarcidae (eelpouts) 23 1 0 . 8 1 87.4 Notothenioidei (includes 5 Antarctic families)d 96 45.1 Tripterygiidae (triplefins) 1 0.5 Achiropsettidae (southern flounders) 4 1.8

Totals 213 100%

"Arranged phylogenetically with sequencing according to Nelson (1 994). bBased on Gon & Heemstra (1990) with notothenioids updated by Eastman & Eakin (2000) and with these additions among non- notothenioids: number of liparids from Andnashev & Stein (1998, p. 58, table 3) plus three from Matallanas (1998, 1999) and number of zoarcids from Anderson (1994). 'Balushkin & Voskoboynikova (1990). dSee Table 111.

278 JOSEPH T. EASTMAN

Liparids, notothenioids and zoarcids are the most speciose fish taxa on the Antarctic shelf. In a large-scale summary (Table I), these groups account for 87.4% of the species. In the most southerly shelf areas (Table II), this figure rises to 91.6%. Such restriction of higher taxonomic diversity is unusual among shelf faunas of the world.

Although liparids are the most speciose family in a large- scale synopsis of the Antarctic Region (Table I), this does not accurately portray aspects of dwersity and biomass. Liparids are not morphologically or ecologically diverse compared with notothenioids. While speciose, many liparid species are known from only a few specimens and are probably not represented by large populations and biomass (Andriashev & Stein 1998). T h s is expected in a group inhabiting the deeper portion of the water column where productivity is reduced. Midwater (DeWitt 1970) and benthic @eWitt 1971, Ekau 1990, Hubold 1992, Eastman & Hubold 1999) trawling in the Ross and Weddell seas indicates that notothenioids always hold overwhelming dominance in species diversity, abundance and biomass - the latter two at levels of 90-95% (Table I1 and references above). T h s is an exceptional degree of habitat saturation by a single higher taxonomic group. Notothenioids are dominant because they occupy niches filled by taxonomically diverse groups offish in temperate and tropical oceans, groups not represented on the Antarctic shelf.

The discrepancy between the large area of the Antarctic shelf and the low species richness may in part be attributable to the absence of certain habitats in the high Antarctic; there are no estuaries, intertidal zones, reefs or shallow shelves. Since the shelfaverages 500 m deep, versus 130 m elsewhere in the world (Kennett 1982), the prime site for fish diversity and abundance is missing in the Antarctic. There is also a phylogenetic constraint that may have limited the size of fauna. Since notothenioids, liparids and zoarcids all lack

Table 11. Fish collected in nineteen bottom trawls at depths of 107-1 191 m on the shelf of the south-westem Ross Sea (73-77"S), cruises 96-6 and 97-9 of the RV Nathaniel B. Palmerb

Percentage by: Family no. of no. of weight

speciesb specimens' (Station 119 only)d

Rajidae 6.4 Muraenolepididae 2.1 Liparidae 6.4 Zoarcidae 8.5 Nototheniidae Artedidraconidae

Bathydraconidae 19.1 Channichthyidae 14.9 Totals 100%

0.7 1.6 0.8 5.3 8.8

76.6 91.6 91.2 18.9 8.1 29.2

100% 100%

'Data from Eastman & Hubold (1999). bNumber of species is 47. 'Number of specimens is 979. '%umber of specimem at Station 119 (77"19'S, 165'41'E; 900-910 m deep) is 83.

swim bladders, they may have been unable to fill all water column niches. In t h s respect, the notothenioid body form has proved to have more capacity for diversification than have those of liparids and zoarcids whlch are keyed to epibenthic and benthic life, respectively. As will be seen below, however, a count of taxa alone leads to an under appreciation of the morphological and ecological diversity of notothenioids.

Notothenioid taxonomy

The suborder Notothenioidei includes eight families, 43 genera with 122 species. As summarized in Table 111, five families and 96 species are Antarctic whereas three families and 26 species are non-Antarctic (Eastman & Eakin 2000). Ninety- three Antarctic notothenioids were included in Fishes of the Southern Ocean (Gon & Heemstra 1990), with an additional 25 non-Antarctic species for a total of 118. The seeming addition of only four species in the past decade is artifactual and d ispses a period of considerable activity in notothenioid taxonomy. While 1 1 new species were described, seven other species were placed in synonymy, hence the net gain of only four (Eastman & Eakm 2000).

Another system of notothenioid taxonomy recognizes more genera and species, approximately 47 and 130, respectively (Balushkin 1992, 1997a, 1997b). Miller (1993), following Balushkin's system, delineated 103 species of Antarctic notothenioids. The difference in taxonomic approach between

Table 111. Species diversity and geographic distribution within notothenioid families and among nototheniid genera.

Antarctic non-Antarctic Total Family" species species species

Bovichtidae Pseudaphritidae Eleginopidaeb Nototheniidae

Patagonotothen Notothenia Paranotothenia Lepidonotothen Gobionotothen Trematomiis Pagothenia Cryothenia Aethotaxis Dissostichiis Gvozdarus Pleuragramma

Harpagiferidae Artedidraconidae Bathydraconidae Channichthyidae

Totals

1 0 0

33 1 3 2 4 4

1 1 2 1 1 2 1 1 6

25 16 15

96

9 1 1

15 13 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0

26

10 1 1

48

6 25 16 15

122

"Familial nomenclature after Balushkin (1992) and phylogenetic sequencing of non-Antarctic families after Lecointre et al. (1997). bDeWitt et al. (1990) rather than Balushkin (1992) is followed for the spelling of the stem-genus portion of the family name Eleginopidae.

ANTARCTIC FISH FAUNA 279

splitting and lumping accounts for the disparity between Balushkin’s and Miller’s counts and that of Gon & Heemstra (1990). The latter is used in this paper and also serves as the basis for the species list of Eastman & Eakin (2000). Most differences between the two systems centre on the families Nototheniidae and Channichthyidae.

The Antarctic fish fauna is far from being completely described. Collecting in accessible and presumably well- stubed areas of the Antarctic shelf is still producing new species. Recently 19 bottom trawls in the Ross Sea yielded 979 specimens including two new artedidraconid species (Eakm & Eastman 1998, Eastman & Eakin 1999), a colour morph of a known arteddraconid (Eastman & Eakin 1999) and two new species of the liparid Paraliparis (Eastman, unpublished data). The new artedidraconids were taken in shallow water (100-300 m) on the tops of banks or in the vicinity of sponge beds. One new liparid was captured less than 50 km offshore, in an innershelf depression at 1200 m, and the other was taken over a bottom that was 465 m deep.

Notothenioid phylogenetics

Much information on phyletic relationshps has appeared in the past few years. This work has utilized DNA sequence data from mitochondria1 genes (Bargelloni et al. 1994, 1997, Ritchie et al. 1996, 1997), nuclear genes (Lecointre et al. 1997) or a combination of both (Bargelloni & Lecointre, 1998, Chen et al. 1998). Amino acid sequences of a- and B- chains of haemoglobin have also been used (Stam et al. 1997, 1998). Notothenioid molecular phylogenetics has provided answers to some questions whilst also opening additional avenues for research. For example, the traditional Bovichtidae was found to be paraphyletic and the newly definedBovichtidae (Bovzchtus and Cottoperca) appears no more closely related to other notothenioids than are several non-notothenioid outgroups (Lecointre et al. 1997). Pseudaphritis urvillii (Valenciennes in Cuvier & Valenciennes), formerly considered a bovichtid, has been reassigned to the monotypic family Pseudaphritidae (Balushkin 1992), andmolecular dataindicate that this species is the sister group of all non-bovichtid notothenioids (Lecointre et al. 1997). Furthermore, it is possible that the families Nototheniidae and Bathydraconidae are paraphyletic and that three offour nototheniid subfamilies are not natural groups (Eiargelloni et al. 1994, Bargelloni & Lecointre 1998, Ritchie et al. 1997, Stam et al. 1997). Obviously much work remains to be done.

The monophyly of notothenioids has not been firmly established on morphologcal grounds. The presence of three plate-like pectoral radials is a possible synapomorphy @akin 1981, Lecointre etal. 1997). The most recent molecular data have not supported monophyly (Lecointre et al. 1997, Bargelloni & Lecointre 1998), but amino acid sequences in haemoglobin suggest that the Antarctic notothenioids are monophyletic (Stam et al. 1997). Lecointre et al. (1997) suggested that a possible reason for this finding is that early

cladogenesis within notothenioids was simultaneous with the explosive diversification of perciforms, thus with little time for differences to accumulate the resolution of outgroup nodes is difficult. Below I discuss the possibility that antifreeze glycopeptides are a synapomorpliy for the clade of Antarctic notothenioids.

Nature of notothenioid biodiversity

Biodiversity is defined as the collection of genomes, species and ecosystems occurring in a geographically defined region (Committee on Biological Diversity in Marine Systems 1995, p. 1). Biodiversity therefore exists at three levels ofbiological organization: genetic (intraspecific), organismal (specific) and ecological (communities). My emphasis here is on organismal biodiversity, although I will also mention intraspecific variation.

One point that should be made initially is that the waters of the Antarctic shelf are not a biodwersity hotspot with large numbers of endemic species found in a relatively small area. Just the opposite is true -a relatively small number of endemic species inhabit a large area. As describedpreviously, it is not the numbers but the nature of the fish biodiversity that dstinguishes the Antarctic from all other shelf areas in the world. The fish fauna lacks the higher taxonomic dwersity characteristic ofmost inshore marine habitats. On the Antarctic shelf, notothenioids dominate the fauna in terms of species diversity, abundance and biomass. Notothenioids have also diversified into a number of life history types - they are morphologically and ecologcally diverse.

Organismal biodiversity: morphological and ecological components

In addition to phyletic &versification, and unlike many other fish groups including some radiations (Brooks & McLennan 1991, p. 171-185, Mayden 1992, p. 870, 876), notothenioids have also experienced substantial morphological and ecological diverslfication. T h s can be summarized by saying that notothenioids underwent a depth-related diversification, most similar to that of the cottoids in Siberian Lake Baikal (Smith &Todd 1984). The notothenioid diversification was keyed to the utilization of unfilled niches in the water column, especially pelagic or partially pelagic zooplanktivory and piscivory. In the absence of competition from a taxonomically diverse fauna, notothenioids were able to fill these niches as well as remaining the dominant benthic group. Their diversification centred on the evolutionary alteration of buoyancy and the morphology associated with swimming and feedmg in the water column. Although they lack swim bladders, in some species density reduction to neutral buoyancy has been achieved through a combination of reduced skeletal mineralization and lipid deposition. Best demonstrated in the family Nototheniidae, about 50% of the Antarctic species inhabit the water column rather than the ancestral benthic habitat (Eastman

280 JOSEPH T. EASTMAN

ANTARCTIC FISH FAUNA 281

1993). Referred to as pelagization, this evolutionary tailoring of morphology for life in the water column has arisen independently in dfferent nototheniid clades (Klingenberg & Ekau 1996).

Figure 1 provides a sample ofthe morphologd, ecological and geographc dlversity among some of the 48 species in the family Nototheniidae (Table 111). While about half of the species are benthic as adults (Fig. la-b, d-g), others are semipelagic (Fig. Ih), epibenthic (Fig. li), cryopelagic (Fig. Ij) and pelagic (Fig. lk-I). With some exceptions, the arrangement of the figure is based on weight in seawater (Eastman 1993) from heaviest at 4.4% (Fig. la) to lightest at 0.6% (Fig. 11). The adults in this sample range in size from 10 cm TL and 30 g (Fig. Id) to 2 m TL and over 100 kg (Fig. Ik). Ecologml diversity in this family is further enhanced by the ontogenetic pelagichenthic habitat shifts, exemplified by juveniles of Notothenia (Fig. lc). Species with complex life cycles contribute extra biodversity to a region (Harper &Hawkworth 1994, p. 10).

Intraspecijc biodiversity

There is another aspect of nototheniid biodiversity that is not conveyed by a count of species. The genus Trematomus, for example, includes species exhibiting variability in head shape, colour and karyology. Although widespread among some species of boreal lacustrine fishes (Skulason & Smith 1995), tlus phenomenon of morphism or phenotypic plasticity has only recently been recognized in Antarctic fishes. I will cite three examples.

The circum-Antarctic nototheniid Trematomus newnesi Boulenger exists as two morphs in the Ross Sea - the typical morph (Fig. lh), which served as the basis for the species description (Boulenger 1902), anda large mouthlbroad headed morph comprising 28% of the population (Eastman & DeVries 1997). These canbe separatedbyvisual inspection. The large mouth morph has a wider and blunter head, longer upperjaw, wider gape and darker coloration. However, a subsequent sample from the same location contained a high percentage (47%) of an intermediate morph (Eastman & DeVries,

unpublished data). Inference from morphology and measurements suggests that the large mouth morph is more benthic than the typical semipelagic morph, although buoyancies have not been measured. This may be an example of diversification along a depth gradient, and its dwovery suggests that the bounds of the nototheniid adaptive radiation may extend to the population level. With the finding of intermediate morphs, other interpretations are also possible. For example, the large mouth morph could be a cryptic or sibling species, and the intermediate morphs are the result of introgression, crosses between large mouth and typical morphs. This hypothesis cannot be tested without data on genetic dwerslfication.

In the only Antarctic example of morphism where genetic information has been obtained, mitochondnal DNA sequences ofthe 12s and 16s ribosomalregions indicatedthat thebrown and white blotch colour morphs of Trematomus bernacchii Boulenger were not genetically different (Bernardi & Goswami 1997). As these mitochondrial regions are conserved, it is possible that genetic differences could have accumulated in more variable regions (Bernardi & Goswami 1997).

Finally, polymorphism in chromosome number and morphology has been discovered in several widely dstributed nototheniids includmg Trematomus eulepidotus Regan and T. hansoni Boulenger (Ozouf-Costaz et al. 1997, Pisano et al. 1998). Although the systematic significanceofthesefindmgs is not clear, intraspecific chromosomal rearrangement may be an initial stage in the attainment of reproductive isolation or it may appear incidentally before or after species formation (Ozouf-Costaz et al. 1997).

The notothenioid speciesjlock

The species flocks of fishes studied to date have lived in freshwater, and most origmated in the past 3 m.y. (million years) (Echelle & Kornfield 1984, McCune 1997). Eastman & Clarke (1998) have drawn parallels between the high Antarctic shelfand lacustrine habitats and, using the criteria of hbbink(1984, p. 22-23), havesuggestedthatthefivehtarctic notothenioid families exhibit the dsproportionate speciosity

Fig. 1. (opposite) Adults and one juvenile (c) depicting the range of morphological, ecological and geographical diversity in the family Nototheniidae. a. Notothenza angustata Hutton, heavy inshore benthic and rockpool species with a cold temperate and sub-Antarctic distribution. b. Adult Notothenza rnicrolepidota Hutton, rocky reef dwelling benthic species with a Subantarctic distribution. c . Pelagic juvenile (86 mm TL) of N. rnzcrolepidota with silvery colouring and forked caudal fin, frequently captured far from shore, d. Patagonotothen eleguns (Giinther), small blenny-like species from southern South America. e. Gobionotothen gbberzfrons (Lonnberg), benthic species common near the Antarctic Peninsula. f. Trematomus bernacchii Boulenger high Antarctic inshore benthic species. g. Trernatomus scotti (Boulenger), high Antarctic offshore benthic species. h. Typical morph of Trematomus newnesi Boulenger, a semipelagic Antarctic species. i. Trematomus loennbergii Regan, epibenthic Antarctic species living near the bottom at depths of 663-1 191 m. j. Pagothenia borchgrevinh (Boulenger), cryopelagic Antarctic species living near the platelet layer underlying fast sea ice. k. Drssostichus eleginoides Smitt, large Subantarctic piscivorous predator probably close to neutral buoyancy and living at depths to 1500 m. 1. Pleuragramma antarcticum Boulenger, shoaling neutrally buoyant zooplanktivore, the dominant species in the water column (0-900 m) of the high Antarctic shelf. [All illustrations, except a-c and g, are of the holotype for the species. Illustrations from these sources: a from Waite (1909); b, c, f, h, j and 1 from Boulenger (1902); d from Gunther ( 1 880); e from Lonnberg (1905); g from Regan (1914), i from Regan (1913) and k from Smitt (1898).]

JOSEPH T. EASTMAN 282

and endemism characteristic of a species flock. Recently another approach has been made to this question

using molecular sequence data. Johns & Avise (1998) attempted to identie phylogenetic footprints characterizing an ancient species flock when viewed at the present time. Using molecular phylogenetic data, they testednull theoretical models that assume random temporal placement of phylogenetic nodes. The null model is rejected if speciation is non-random and if there is significant clustering of cladogenetic events in time. Their analysis detected an ancient marine species flock, with an estimated age 8.5-3.6 m.y., among the 65 species of north-eastern Pacific scorpaeniform rockfishes of the genus Sebastes. Johns & Avise (1998) conducted a similar analysis using the notothenioid molecular data of Bargelloni et al. (1994) and Ritchieetal. (1996). Whenbovichtids wereexcluded, statistical tests of nodal placements in the notothenioid phylogeny suggested that speciation events were non-random, with signifcant clustering of cladogenetic events in time - the hallmark of a species flock. This was true for the Antarctic notothenioid families in general and also for the nototheniid trematomids in particular. The dates of these two rahations hadbeenpreviously estimated at 10-15 m.y. (Bargelloni etal. 1994) and 3.4 m.y. mtchie et al. 1996), respectively.

A key innovation

A key innovation is a uniquely derived novel feature, or synapomorphy, characteristic of a clade and correlated with theradiationofthat clade(Brooks&McLennan 1991 ,~ . 181). The innovation must also aid entry to a new ecological habitat (Grant 1998, p. 3 13-3 14). Key innovations are the exception rather than the rule among flocks of fishes. The five different varieties of peptide and glycopeptide antifreezes have evolved from pre-existing proteins by gene duplication followed by sequence hvergence (Davies et al. 1993, Chen et al. 1997a, 1997b, Cheng 1998a). This has occurred independently at least nine times among phyletically diverse cold water teleosts (Davies et al. 1993, Cheng 1998b). Antifreeze glycopeptides ofhtarcticnotothenioids meet the criteria for a key innovation. When mapped on the notothenioid cladogram, they have a single origin at the node for the Antarctic clade - the sister group of the non-Antarctic Eleginopidae (Eastman & Clarke 1998). Given their orign from a transformed trypsinogen gene, they are uniquely derived and distinct from all other antifreezes including the antifreeze glycopeptides of Arctic gadids which have a different but unidentified genomic origin (Chen et al. 1997a, 1997b, Cheng 1998a). The contribution of antifreeze glycopeptides to survival in new habitat is obvious since diversification into subzero ice-laden water would have been impossible without them @eVries 1988). There has been considerably more Qverslfication in the Antarctic component of the suborder (95 species) than in the non-Antarctic (26 species). Based on the 14-5 m.y. age inferred from the divergence of molecular sequences between

trypsinogenandantifreeze (Chen etaE. 1997a), the acquisition of genes responsible for the synthesis of antifreeze in the nototheniid Dissostichusmawsoni Norman is coincident with the cooling and appearance of ice in the Southern Ocean.

Radiations within the radiation

Repeated diversification into water column habitats has been the hallmark of the notothenioid radiation. Althoughthe most thoroughly studied cases are in the family Nototheniidae (DeVries & Eastman 1978, Eastman & DeVries 1981,1982, Klingenberg & Ekau 1996), there may also be examples among icefish of the family Channichthyidae. A molecular phylogeny is available for channichthyids and mapping of lifestyle data suggests there have been benthic-pelagic &versifications at most nodes in the cladogram (Chen et al. 1998). This hypothesis is based on existing life history information, whch is scant for some species, and buoyancies have not been measured. Some species do have a morphology associated with habitation of the water column. Dacodraco hunteri Waite, for example, has a weakly ossified skeleton with considerable cartilage in the skull (Eastman 1999). It has a partially persistent notochord and reduced amounts of bone in the vertebral column since the centra are incompletely constricted. Since their weight in seawater averages only 1.3% of weight in air, Dacodraco are among the lightest notothenioids. It is likely that other channichthyids possess similar morphology since heterochrony is thought to have been important in the &versification of several notothenioid lineages into water column habitats (Balushkin 1984, p. 127-128, Iwami 1994, Voskoboinikova 1994). Delayed ossification is a paedoniorphic trait, most pronounced in the phyletically derived channichthyid species (Voskoboinikova 1997).

The fossil record and the evolution of notothenioids

Does information from the fossil record contribute to understanhng the nature of modern Antarctic fish fauna, with greatly restricted higher taxonomic diversity unlike that of any other shelf area in the world? Ths question is an attempt to understand a faunal replacement that took place sometime over the past 40 m.y. in the waters of the Gondwanan shelf. The null hypothesis is that there would be some taxonomic carryover from the Eocene to the modern Antarctic fauna, as was true in the case ofother southern continents like Australia (Long 1982). The Cenozoic fossil record provides a starting point, but beyond this speculation predominates.

Eocene fauna

As is the case with many extant groups, there is no fossil fauna that is immediately ancestral to the modern Antarctic fish fauna. There is, however, a late Eocene (-40 Ma) fauna. These fossil fish from the La Meseta Formation on Seymour

ANTARCTIC FISH FAUNA 283

Island, near the tip ofthe Antarctic Peninsula, are representative of a high latitude (z6OoS) shelffauna at this time. Seymour Island had a taxonomically diverse fish fauna that was cool temperate in character (Eastman 1993). As biogeographic provinciality was less evident at high latitudes in the Eocene, this fauna was cosmopolitan rather than endemic to the Antarctic component of Gondwana.

Chondrichthyes - cartilaginousflshes

The Eocene fish fauna from Seymour Island consists of about 26 taxa-20 chondrichthyans (Case 1992, Long 1992a, Cione & Reguero 1998) and 6-7 actinopterygians (Eastman 1993, Long 1992b, Cioneetd. 1995). Therangeofchondrichthyans includes chimaeras, large predatory sharks, sedentary benthlc sharks, saw sharks, skates and eagle rays. However the 70% dominance by chondrichthyans is artifactual, attributable to the durability and distinctiveness of their dentition in the fossil record. Their teeth are frequently identfiable to species, but this is not the case with most actinopterygian remains from SeymourIslandor anywhere else (Grandelk Chatterjee 1987). Chondrichthyans probably composed a greater percentage of the Eocene fauna than the 4% in the modern Antarctic fauna (Gon&Heemstra 1990), but theexact figure isuncertain. The majority of chondrichthyan species in the modern fauna are skates of the genera Bathyraja and Raja. The modern cool temperate Tasmanian fauna includes 15% chondrichthyans (Last et al. 1983). Case (1992) suggests that during the late Eocene the east coast of the Antarctic Peninsula was an area of upwelling and therefore capable of supporting a high biomass of prey necessary for maintenance of a diverse shark fauna. The recent report of a fossilized gill raker from the basking shark Cetorhinus (Cione & Reguero 1998), a specialized filter feeder on zooplankton, supports this hypothesis. The filter feeding strategy is not represented among chondrichthyans or actinopterygians in the modern Antarctic fauna.

Isit possible that skates have continuously occupied Antarctic coastal waters since the early Tertiary, or are they Quaternary immigrants? Skates of the genera Bathyraja and Raja are the only fishes with familial and generic representation inboth the Eocene and modern faunas (Long 1994). Although favouring persistence, Long (1994) objectively summarizes both sides of the issue. This question should be answerable using molecular techniques to infer phylogenetic relationships and divergence times among rajids of the Southern Hemisphere.

Actinopterygii - ray-flnnedjshes

The actinopterygian fossils from Seymour Island include clupeoids (herrings), silunforms (catfish), gaMorms (codfish), oplegnathids (knifejaws), labrids (wrasses) and trichiurids (cutlassfish). The teeth and durophagous diet of oplegnathids and labrids suggest hard shelled invertebrate prey was then abundant. With the exception of gadiforms, these taxa are not

represented in the modem Antarctic fauna. The representation of notothenioids in the fossil fish fauna from Seymour Island is debatable. A partial cranium has been identified as either a gadiform (Eastman & Grande 199 1) or a representative of the basal notothenioid family Eleginopidae, Proeleginops grandeastmanorum (Balushkin 1994).

With reference to the perciform fossil record elsewhere in the world, 40 million years is a reasonable date for the possible existence of a basal notothenioid group on the Gondwanan sheK. There are no indisputable perciform fossils in Cretaceous deposits, but by the early Middle Eocene (z50 Ma) most perciform families were represented to the extent that subsequent diverslfication was “mere tinkering” (Patterson 1993, p. 53). For example, theEocene Monte Bolca coral reef assemblage from northern Italy, deposited in what was then the Tethys Sea, is broadly similar in familial composition to modern reef assemblages (Bellwood 1996). This suggests that the evolution of perciforms was rapid, with basic body plans and familial lineages established by 50 Ma followed by a period of stasis. Perciforms dominated shallow tropical marine communities at this time. It is not surprising that the versatile perciform body plan was also suitable for benthic habitats in high southern latitudes.

Pliocene fauna

With little ice-free rock exposed in Antarctica, there is scant information about any transitional or intermedate faunas spanning the gap between the late Eocene and modern faunas. Pliocene (4.2-3.5 Ma) invertebrate and cetacean fossils from Marine Plain in the Vestfold Hills indicate that the inshore fauna, which also includes one unidentlfiable fish, was unlike the modern fauna (Quilty 1994). For example, decapod crustaceans have a long fossil history in Antarctica but are unknown in the modern fauna. A Pliocene decapod from the Marine Plain verifies that this group was still represented in coastal waters at this time (Feldmann & Quilty 1997), as the result of either persistence or re-invasion. The diversity of cetaceans from the Marine Plain deposits is also noteworthy and unlike the modern coastal fauna. Five species have been discovered to date, includmg an 8.5 m-long squid-eating dolphin (Quilty 1994).

The Cenozoic paleoenvironment and the divers$cation of notothenioids

During the early Miocene (25-22 Ma) the Antarctic shelf was subject to a series of tectonic and oceanographic events that probably altered faunal composition. For example, seafloor spreading and the opening of the Drake Passage to deep water began to separate Antarctica from other Gondwanan landmasses (Kennett 1982). The development of the unrestricted Antarctic Circumpolar Current and initial formation ofthe Antarctic Polar Front, with increasing biogenic productivity and somewhat lower temperatures to the south

284 JOSEPH T. EASTMAN

sennett 1982), began to define the Southern Ocean and Antarctic Zoogeographic Region. These events isolated Antarctica and initiated climatic changes that eventually eliminated some habitats available to the late Eocene Seymour Island fish fauna. After the middle Miocene (016 Ma), expansion of the ice sheet led to destruction of inshore habitat by ice, with repeated grounlngs of parts of the sheet as far as the shelf break (Anderson 1999, p. 229-230). Invertebrate faunas and trophic conltions changed as well. Emigration, regional extirpation and/or extinction reduced the diversity of the fish fauna, and its taxonomic composition was altered through geologic time. There was a post-Eocene faunal replacement with little carryover of families into the modern fauna.

Diversification of notothenioids was probably facilitated by the thermal isolation of Antarctica, by the increasing productivity of the Southern Ocean beginning about 22 Ma (Kennett 1982) and by the absence of competition from non- notothenioids. Niches were available and, as a generalized benthic group, notothenioids were able to adapt to new conditions. There was also probably an element of chance in the emergence of the notothenioid stock - being in the right place at the right time. The notothenioid diversification may have been protracted through much of the Tertiary, with cladogenetic events separated by long periods of stasis. For example, the divergence of some of the basal non-Antarctic clades, the Pseudaphritidae in Australia and the Eleginopidae in South America, may have taken place by vicariance during the fragmentation of Gondwana when stocks became established on the coasts of isolated continental blocks (Eastman 1993). The separation of the eleginopid clade, represented by the extant Eleginops maclovinus (Cuvier in Cuvier & Valenciennes), from the rest of the non-bovichtid notothenioids is estimated at 23-22 m.y. (Bargelloni & Lecointre 1998). Ths coincides with the opening oftheDrake Passage and formation of the Antarctic Polar Front.

The phyletically derived Antarctic clades appeared later. Inferred dates for the age of the radiation of the five Antarctic notothenioid families, those with antifreeze glycopeptides, are in the range of 15-5 m.y. For example, inference from mitochondria1 DNA sequences implies that the radiation of this group took place within the last 15 m.y. (Bargelloni et al. 1994) or 16-10 m.y. (Bargelloni & Lecointre 1998). The small sequence divergence between notothenioid antifreeze glycopeptide and trypsinogen genes provides a similar age, 14-5 m.y., for the radiation of the Antarctic families. The inferred date for this gene conversion encompasses the time of appearance of ice in the Southern Ocean (Chen et al. 1997a). In the family Nototheniidae, the average age of the trematoniid radiation is only 3.4 m.y. or late Pliocene (ktchie et al. 1996). Trematomids are lugh latitude inshore fish with high levels of antifreeze. Although controversial, the late Pliocene (3.5-1.8 Ma) may have been a time of climatic and glacial instability at high Antarctic latitudes (Quilty 1996). If Antarctica was partially deglaciated and a rise in sea level

increased the area of the shelf, these warm-dominatedvicarant events could also have provided new habitat suitable for diversification of some notothenioids (Eastman 1993,

Finally, why is the modern Antarctic faunadistinctly different from those of other southern continents when all shared a common early Tertiary fish fauna? The answer probably lies in the regional historical processes mentioned previously, especially tectonic, oceanographic and climatic events. These processes are important in establishing patterns of diversity in polar marine organisms (Brey et al. 1994). With the breakup of Gondwana, each southern landmass and its shelf fauna became independent and influenced by lfferent hstorical processes. As the Tertiary progressed, Antarctic shelf waters became increasingly isolated and subject to current patterns, glacial scour, changing temperatures, marked seasonality and altered trophic conditions. By the late Miocene-late Pliocene (6.5-2 Ma), low water temperatures, extreme geographc isolation and lack of south flowing surface currents limited immigration of many Southern Hemisphere epipelagic and inesopelagic groups into waters south of the Antarctic Polar Front. The shelf fauna became increasingly endemic and distinctive, but it is not known when it became modern in taxonomic composition.

p. 143-144).

Future research

The development of Antarctic evolutionary biology has benefited greatlyfrom work on notothenioidfish. IfAntarctica is to gainvisibility as adistinctiveevolutionary site, additional work on fish in this unique ecosystem will be necessary. Ship- based collecting as well as basic taxonomic and ecological research should continue. The shelf waters are still yielding new species and presenting challenging problems, such as intraspecific variation and morphisni, in the delimitation of species. The application of phylogenetic systematics has allowed formulation of testable hypotheses of historical descent among notothenioid lineages. This methodology also provides a framework for tracing character acquisition during diversification and thus opens up new approaches to longstanding evolutionary mysteries. The most valuable research in this area has answered broad organismal questions by employing the techniques of molecular biology to track the acquisition of a key innovation via gene conversion (Chen et al. 1997a, 1997b, Cheng 1998a). The questions in this case: where did notothenioid antifreezes come from, when did they appear and do they have aunique origin among fish antifreezes? The surprisingly direct answers were mentioned above in the section on key innovations. It is hoped that work along these lines will continue.

The exploration of remote islands has proved to be far more valuable than merely documenting the existence of unusual faunas. Discoveries made at the Galapagos stimulated Charles Darwin to begin erecting the framework for evolutionary thought. Darwin’svoyageintheHMS Beagle(l83 1-36), and

ANTARCTIC FISH FAUNA 285

subsequent research by numerous scientists, has made the Galapagos the premier evolutionary site in the world. At about the same time (1839-43), James Clark Ross took HMS Erebus and Terror into the h g h latitudes of Antarctic coastal waters (Ross 1982). During the course of the voyage many distinctive species were discovered including an endemic tribe of four seals, six species of penguins and about a dozen species of notothenioid fish, nine of which are still valid. However, a comprehensive picture of the fauna did not emerge and Antarctica remained unappreciated as a continental island with an endemic marine fauna. Antarctica has now joined the Galapagos as a destination of tourist ships and fishingvessels, but unlike the Galapagos, it still does not have name recognition as a centre of evolution. This SCAR Workshop inQcates that scientists are now tapping the potential of the Antarctic for evolutionary studies.

Acknowledgements

I thank Bruno Battaglia and the SCAR Subcommittee on the Evolutionary Biology ofAntarctic Organisms for the invitation to speak at the workshop. I am also grateful to Edith Fanta for her efforts in arranging the meeting and for her hospitality in Curitiba, Brazil. The manuscript benefited greatly from the reviews of Richard Eakin and Ofer Gon. Supported by NSF grant OPP94-16870 and by funds from an Ohio University Presidential Research Scholar Award.

References

ANDERSON, J.B. 1999. Antarctic marine geology. Cambridge: Cambridge University Press, 289 pp.

ANDERSON, M.E. 1994. Systematics and osteology of the Zoarcidae (Teleostei: Perciformes). Ichthyological Bulletin, J.L.B. Smith Institute of Ichthyology, No. 60, 1-120.

ANDRIASHEV, A.P. 1965. A general review ofthe Antarctic fish fauna. In VAN OYE, P. & V A N MIEGHEM, J., eds. Biogeography and ecology in Antarctica, Monographiae biologrcae, vol. XV. The Hague: Junk, 491-550.

ANDRIASHEV, A.P. 1986. Review of the snailfish genus Paraliparis (Scorpaeniformes: Liparididae) ofthe Southern Ocean. Koenigstein: Koeltz Scientific Books, Theses Zoologicae, vol. 7, 204 pp.

ANDRIASHEV, A.P. 1987. A general review of the Antarctic bottom fish fauna. In KULLANDER, S.O. & FERNHOLM, B., eds. Fijlh congress of European ichthyologists, proceedings, Stockholm. I985. Stockholm: Swedish Museum of Natural History, 357-372.

ANDRIASHEV, A.P. & STEIN, D.L. 1998. Review of the snailfish genus Careproctus (Liparidae, Scorpaeniformes) in Antarctic and adjacent waters. Contributions in Science. Natural History Museum of Los Angeles County, No. 470, 1-63.

BALUSHKIN, A.V. 1984. Morphological bases of the systematics and phyfogeny ofthe nototheniidfishes. Published for Division of Polar Programs, National Science Foundation by Amerind Publishing Co., New Delhi, India, 1989, 153 pp. [Translated from Russian.]

BALUSHKIN, A.V. 1992. Classification, phylogenetic relationships, and origins of the families of the suborder Notothenioidei (Perciformes). Joiirnal of Ichthyology, 32(7), 90-1 10.

BALUSHKIN, A.V. 1994. Proeleginops grandeastmanorum gen. et sp. nov. (Perciformes, Notothenioidei, Eleginopsidae) from the Late Eocene of Seymour Island (Antarctica) is a fossil notothenioid, not a gadiform. Journal of Ichthyology, 34, 10-23.

BALUSHKIN, A.V. 1997a. Morfologiya, klassifikatsiya I evolyutsiya nototenioidnykh ryb Yurhnogo Okeana. [Morphology, Classification, and Evolution of Notothenioid Fishes of the Southern Ocean]. St Petersburg: Doctor of Science Dissertation, Zoological Institute, Russian Academy of Sciences, 52 pp. [In Russian.]

BALUSHKIN, A.V. 1997b. Recent knowledge onthe system and phylogeny of the notothenioid fishes. European Science Foundation Network Fishes of the Antarctic Ocean, Final Conference: Evolution of the Antarctic Ichthyofauna. Pontignano, Siena, Italy, 22-24May 1997. [Abstract.]

BALUSHKIN, A.V. & VOSKOBOYNIKOVA, 0 , s . 1990. A new family, Bathylutichthyidae (Cottoidei, Scorpaeniformes), for the deepwater fish Bathyltitichthys taranetzi gen. et sp. nov. from South Georgia Island (Antarctica). Journal of Ichthyology, 30, 67-75.

BARGELLONI, L. & LECOINTRE, G. 1998. Four years in notothenioid systematics: a molecular perspective. In DI PRISCO, G., PISANO, E. & CLARKE, A., eds. Fishes of Antarctica: a bioiogical overview. Milan: Springer Verlag, 259-273.

BARGELLONI, L., PATARNELLO, T., RITCHIE, P.A., BATTAGLIA, B. & MEYER, A. 1997. Molecular phylogeny and evolution of notothenioid fish based on partial sequences of 1 2 s and 1 6 s ribosomal RNA mitochondrial genes. In BATTAGLIA, B., VALENCIA, J. & WALTON, D. W.H., eds. Antarctic communities: species, structure andsurvival. Cambridge: Cambridge University Press, 45-50.

BARGELLONI, L., RITCHIE, P. A., PATARNELLO, T., BATTAGLIA, B., LAMBERT, D.M. & MEYER, A. 1994. Molecular evolution at subzero temperatures: mitochondrial and nuclear phylogenies of fishes from Antarctica (suborder Notothenioidei), and the evolution of antifreeze glycopeptides. Molecular Biology and Evolution. 11, 854-863.

BELLWOOD, D.R. 1996. The Eocene fishes of Monte Bolca: the earliest coral reef fish assemblage. Coral Reefs, 15, 11-19.

BERNARDI, G. & GOSWAMI, U. 1997. Molecular evidence for cryptic species among the Antarctic fish Trematomus bernacchii and Trematomiis hansoni. Antarctic Science. 9, 381-385.

BOULENGER, G.A. 1902. Pisces In Report on the collections of natitral history made in the Antarctic regions during the voyage of the “Southern Cross ”. London: British Museum (Natural History), 174-189, plates XI-XVIII.

BREY, T., KLAGES, M., DAHM, C., GORNY, M., GUTT, J., HAIN, S. , STILLER, M., ARNTZ, W.E., WAGELE, J.-W &ZIMMERMANN, A. 1994. Antarctic benthic diversity. Nature. 368, 297.

BROOKS, D.R. & MCLENNAN, D.A. 1991. Phylogeny, ecology, and behavior: a research program in comparative biology. Chicago: University of Chicago Press, 434 pp.

CASE, J.A. 1992. Evidence from fossil vertebrates for a rich Eocene Antarctic marine environment. Antarctic Research Series, 56, 119-130.

CHEN, W.-J., BONILLO, C. & LECOINTRE, G. 1998. Phylogeny of the Channichthyidae (Notothenioidei, Teleostei) based on two mitochondrial genes. In DI Fkrsco, G., PISANO, E. & CLARKE, A , eds. Fishes ofAntarctica: a biological overview. Milan: Springer Verlag, 287-298.

CHEN, L., DEVRIES, A.L. & CHENG, C.-H.C. 1997a. Evolution of antifreeze glycoprotein gene from a trypsinogen gene in Antarctic notothenioid fish. Proceedings of the NationalAcademy ofSciences of the USA, 94, 3811-3816.

CHEN, L., DEVRIES, A.L. & CHENG, C.-H.C. 1997b. Convergent evolution of antifreeze glycoproteins in Antarctic notothenioid fish and Arctic cod. Proceedings of the National Academy of Sciences of the USA, 94, 3817-3822.

286 JOSEPH T. EASTMAN

CHENG, C.-H.C. 1998a. Origin andmechanism ofevolution of antifreeze glycoproteins in polar fishes. In DI PRISCO, G., PISANO, E. & CLARKE, A., eds. Fishes of Antarctica: a biological overview. Milan: Springer Verlag, 31 1-328.

CHENG, C.-H.C. 1998b. Evolution of the diverse antifreeze proteins. Current Opinion in Genetics & Development, 8, 715-720.

CIONE, A.L. & REGUERO, M.A. 1998. A middle Eocene basking shark (Lamniformes, Cetorhinidae) from Antarctica. Anfarctic Science, 10, 83-88.

CIONE, A.L., DE LAS MERCEDES AZPELICUETA, M. & BELLWOOD, D.R. 1995. An oplegnathid fish from the Eocene of Antarctica. Palaeontology, 37, 93 1-940.

CLARKE, A. & JOHNSTON, LA. 1996. Evolution and adaptive radiation of Antarctic fishes. Trends in Ecology & Evolution, 11, 212-218.

COMMITTEE ON BIOLOGICAL DIVERSITY IN MARINE SYSTEMS. 1995. Understanding marine biodiversity: a research agenda for the nation. Washington: National Academy Press, 114 pp.

DAVIES, P.L., EWART, K.V. & FLETCHER, G.L. 1993. The diversity and distribution offish antifreeze proteins: new insights into their origins. In HOCHACHKA, P.W. & MOMMSEN, T.P., eds. Biochemistry and molecular biology of fishes. vol. 2, Molecular biology frontiers. Amsterdam: Elsevier, 279-291.

DEVRIES, A.L. 1988. The role of antifreeze glycopeptides and peptides in the freezing avoidance of Antarctic fishes. Comparative Biochemistry and Physiology, 90B, 6 1 1-62 1.

DEVRIES, A.L. & EASTMAN, J.T. 1978. Lipid sacs as a buoyancy adaptation in an Antarctic fish. Nature, 271, 352-353.

DEWITT, H.H. 1970. The character of the midwater fish fauna of the Ross Sea, Antarctica. In HOLDGATE, M. W., ed. Antarctic ecology, vol. 1 . London: Academic Press, 305-3 14.

DEWITT, H.H. 1971. Coastal and deep-water benthic fishes of the Antarctic. In BUSHNELL, V.C., ed. AntarcticMap Folio Series, Folio 15. New York: American Geographical Society, 1-10,

DEWITT, H.H., HEEMSTRA, P.C. & GON, 0. 1990. Nototheniidae. In GON, 0. & HEEMSTRA, P.C., eds. Fishes of the Southern Ocean. Grahamstown, South Africa: J.L.B. Smith Institute of Ichthyology, 279-331.

DI PRISCO, G. 1998. Molecular adaptations ofAntarctic fish hemoglobins. In DI PRISCO, G., PISANO, E. & CLARKE, A., eds. Fishes ofAntarcticat a biological overview. Milan: Springer Verlag, 339-353.

DI PRISCO, G., MARESCA, B. & TOTA, B., eds. 1991. Biology ofAntarctic fish. Berlin: Springer Verlag, 292 pp.

DIPRISCO, G., PISANO, E. &CLARKE, A., eds. 1998. Fishes ofAntarctica: a biological overview. Milan: Springer Verlag, 363 pp.

EAKIN, R.R. 1981. Osteology and relationships of the fishes of the Antarctic family Harpagiferidae (Pisces, Notothenioidei). Antarctic Research Series, 31, 81-147.

EAKIN, R.R. & EASTMAN, J.T. 1998. New species of Pogonophryne (Pisces, Artedidraconidae) from the Ross Sea, Antarctica. Copeia, No. 4, 1005-1009.

EASTMAN, J.T. 1993. Antarctic fish biology: evolution in a iiniqzie environment. San Diego: Academic Press, 322 pp.

EASTMAN, J.T. 1999. Aspects of the biology of the icefish Dacodraco hunteri (Notothenioidei, Channichthyidae) in the Ross Sea, Antarctica. Polar Biology. 21, 194-196.

EASTMAN, J.T. & CLARKE, A. 1998. Radiations of Antarctic and non- Antarctic fish. In D I PRISCO, G., PISANO, E. & CLARKE, A, , eds. Fishes ofAntarctica: a biological overview. Milan: Springer Verlag, 3-26.

EASTMAN, J.T. & DEVRIES, A.L. 1981. Buoyancyadaptations inaswim- bladderless Antarctic fish. Journal of Morphology, 167, 91-102.

EASTMAN, J.T. & DEVRIES, A.L. 1982. Buoyancy studies ofnotothenioid fishes in McMurdo Sound, Antarctica. Copeia, No. 2, 385-393.

EASTMAN, J.T. & DEVRIES, A.L. 1997. Biology and phenotypic plasticity of the Antarctic nototheniid fish Trematomiis newnesi in McMurdo Sound. Antarctic Science, 9, 27-35.

EASTMAN, J.T. & EAKIN, R.R. 1999. Fishes of the genus Artedidraco (Pisces, Artedidraconidae) from the Ross Sea, Antarctica, with the description of a new species and a colour morph. Antarctrc Science, 11, 13-22.

2000. An updated species list for notothenioid fish (Perciformes; Notothenioidei), with comments on Antarctic species. Archive of Fishery and Marine Research, 48, 11-20.

EASTMAN, J.T. & GRANDE, L. 1991. Late Eocene gadiform (Teleostei) skull from Seymour Island, Antarctic Peninsula. Antarctic Science, 3, 87-95.

EASTMAN, J.T. & HUBOLD, G. 1999. The fish fauna of the Ross Sea, Antarctica. Antarctic Science, 11, 293-304.

ECHELLE, A.A. & KORNFIELD, I., eds. 1984. Evolution offish species flocks. Orono, ME: University of Maine at Orono Press, 257 pp.

EKAU, W. 1990. Demersal fish fauna of the Weddell Sea, Antarctica. Antarctic Science, 2, 129-137.

ESCHMEYER, W.N., ed. 1998. Catalog offishes, vol. 1 . San Francisco: California Academy of Sciences, 958 pp,

FELDMANN, R.M. & QUILTY, P.G 1997. First Pliocene decapod crustacean (Malacostraca: Palinuridae) from the Antarctic. Antarctic Science, 9, 56-60.

GIVNISH, T.J. 1997. Adaptive radiation and molecular systematics: Issues and approaches. In GIVNISH, T. J. & SYTSMA, K. J., eds. Molecular evoltition andadapfive radrafion. Cambridge: Cambridge University Press, 1-54.

GON, 0. & HEEMSTRA, P.C., eds. 1990. Fishes of the Southern Ocean. Grahamstown, South Africa: J.L.B. Smith Institute of Ichthyology, 462 pp.

Seymour Island, Antarctic Peninsula. Palaeontology, 30, 829-837. GRANT, P.R., ed. 1998. Evoltition on islands Oxford: Oxford

University Press, 334 pp, GROVE, J.S. & LAVENBERG, R.J. 1997. Thefishes of the Galapagos

Islands. Stanford: Stanford University Press, 863 pp. GUNTHER, A. 1880. Report on the shore fishes procured during the

voyage 0fH.M.S. Challenger in the years 1873-1876. Report on the Scientific Results of the Voyage of H.M.S. Challenger during the years 1873-1876, Zoology, 1(6), 1-82, plates VIII-XI.

HARPER, J.L. & HAWKSWORTH, D.L. 1994. Biodiversity: measurement and estimation. Philosophical Transactions of the Royal Societv, London. 345B, 5-12.

HELFMAN, G.S., COLLETTE, B.B. & FACEY, D.E. 1997. The diversity of fishes. Malden, MA: Blackwell Science, 528 pp.

HUBOLD, G. 1992. Zurokologie der Fische im Weddellmeer. Berichte ziw Polarforschting, 103, 1-1 57. [English summary.]

HUBOLD, G. & EKAU, W. 1987. Midwater fish fauna ofthe Weddell Sea, Antarctica. In KULLANDER, S . 0 & FERNHOLM, B., eds. Fifrh congress of Enropean ichthyologists, proceedings, Stockholm, 1985. Stockholm: Swedish Museum of Natural History, 391-396.

IWAMI, T. 1994. Heterochrony in the family Channichthyidae (Perciformes, Notothenioidei) and its phylogenetic implications. Sixth SCAR Symposiiim on Antarctic Biology, Venice, Italy, 30 May-3 June 1994, p. 134. [Abstract.]

JOHNS, G.C. & AVISE, J.C. 1998. Tests for ancient species flocks based on molecular phylogenetic appraisals of Sebasfes rockfishes and other marine fishes. Evolution, 52, 1135-1146.

KENNETT, J.P. 1982. Marinegeology. Englewood Cliffs, NJ: Prentice- Hall, 813 pp.

KLINGENBERG, C.P. & EKAU, W. 1996. A combined morphometric and phylogenetic analysis of an ecomorphological trend: pelagization in Antarctic fishes (Perciformes: Nototheniidae). Biological Jotirnal of the Linnean Society, 59, 143-177.

KOCK, I<.-H. 1992. Antarcticfish andfisheries. Cambridge: Cambridge University Press, 359 pp.

EASTMAN, J.T. & EAKIN, R.R.

GRANDE, L. & CHATTERIEE, s. 1987. New Cretaceous fish fossils from

ANTARCTIC FISH FAUNA 287

LAST, P.R., SCOTT, E.O.G. & TALBOT, F.H. 1983. Fishes of Tasmania. Hobart: Tasmanian Fisheries Development Authority, 563 pp.

LECOINTRE, G., BONILLO, C., OZOUF-COSTAZ, C. & HUREAU, J.-C. 1997. Molecular evidence for the origins of Antarctic fishes: paraphyly of the Bovichtidae and no indication for the monophyly of the Notothenioidei (Teleostei). Polar Biology, 18, 193-208.

,ONG, D.J. 1992a. Sharks from the La Meseta Formation (Eocene), Seymour Island, Antarctic Peninsula. Journal of Vertebrate Paleontology, 12, 11-32.

(ONG, D.J. 1992b. An Eocene wrasse (Perciformes; Labridae) from Seymour Island. Antarctic Science. 4, 235-237.

ONG, D. J. 1994. Quaternary colonization or Paleogene persistence?: historical biogeography of skates (Chondrichthyes: Rajidae) in the Antarctic ichthyofauna. Paleobiology, 20, 21 5-228.

LONG, J. 1982. The history of fishes on the Australian continent. In RICH, P.V. & THOMPSON, E.M., eds. The fossil vertebrate record of Australasia. Clayton: Monash University Offset Printing Unit, 53-85.

LONNBERG, E. 1905. The fishes ofthe Swedish South Polar Expedition. Wissenschaftliche Ergebnisse der Schwedischen Siidpolar- Expedition 1901-1903. 5, 1-72, plates I-V.

MATALLANAS, J. 1998. Description of Careproctus guillemr n sp. (Pisces: Scorpaeniformes) from the Weddell Sea. Journal of Fish Biology, 52, 380-385.

MATALLANAS, J. 1999. New and rare snailfish genus Paraliparis from the Weddell Sea with the description of two new species. Joirrnal of

MAYDEN, R.L. 1992. An emerging revolution in comparative biology and the evolution of North American freshwater fishes. In MAYDEN, R.L., ed. Systematics, historical ecology, and North American freshwater fishes. Stanford: Stanford University Press, 864-890.

MCCOSKER, J.E. 1998. Book review of Thefishes of the Galapagos Islands by J.S. Grove & R.J. Lavenberg. Copeia, No. 3, 809-812.

MCCUNE, A.R. 1997. How fast is speciation? Molecular, geological, and phylogenetic evidence from adaptive radiations of fishes. In GIVNISH, T. J. & SYTSMA, K. J., eds. Moleciilar evolution andadaptive radiation. Cambridge: Cambridge University Press, 585-6 10.

MILLER, R.G. 1993. History and atlas of the fishes of the Antarctic Ocean. Carson City, NV: Foresta Institute for Ocean and Mountain Studies, 792 pp.

NELSON, J.S. 1994. Fishes ofthe world(3rd edition). New York: John Wiley & Sons, 600 pp.

OZOUF-COSTAZ, C., PISANO, E., THAERON, C. & HUREAU, J.-C. 1997. Antarctic fish chromosome banding: significance for evolutionary studies. Cybium, 21, 399-409.

PAKHOMOV, E.A. 1997. Feeding and exploitation ofthe food supply by demersal fishes in the Antarctic part of the Indian Ocean. Joitrnal of Ichthyology, 37, 360-380.

PATTERSON, C. An overview of the early fossil record of acanthomorphs. Bulletin of Marine Science, 52, 29-59.

PISANO, E., OZOUF-COSTAZ, C. & PRIRODINA, V. 1998. Chromosome diversification in Antarctic fish (Notothenioidei). In DI PRISCO, G., PISANO, E. & CLARKE, A., eds. Fishes of Antarctica: a biological overview. Milan: Springer Verlag, 275-285.

Fish Biology, 54, 1017-1028.

1993.

QUILTY, P.G. 1994. Preliminary results of 1993/1994 fieldwork, Marine Plain, Vestfold Hills, East Antarctica. Terra Antartica, 1,

QUILTY, P.G. 1996. The Pliocene environment of Antarctica. Papers and Proceedings ofthe Royal Society of Tasmania, 130, 1-8.

REGAN, C.T. 1913. The Antarctic fishes of the Scottish National Antarctic Expedition. Transactions of the Royal Society of Edinburgh, 49(2), 229-292, plates I-XI.

REGAN, C.T. 1914. Fishes. BritishAntarctic (“TerraNova”) Expedition, 1910, Natural History Report, Zoology, 1, 1-54, plates I-XIII.

RIBBINK, A.J. 1984. Is the species flock concept tenable? In ECHELLE, A.A. & KORNFIELD, I., eds. Evolution off ish speciesflocks. Orono: University of Maine at Orono Press, 21-25.

RITCHIE, P., LAVOU~, S. & LECOINTRE, G. 1997. Molecularphylogenetics and the evolution of Antarctic notothenioid fishes. Comparative Biochemistry and Physiology, 118A, 1009-1025.

RITCHIE, P.A., BARGELLONI, L., MEYER, A., TAYLOR, J.A., MACDONALD, J.A. & LAMBERT, D.M. 1996. Mitochondria1 phylogeny oftrematomid fishes (Nototheniidae, Perciformes) and the evolution of Antarctic fish. Molecular Phylogenetics and Evolution, 5, 383-390.

Ross, M.J. 1982. Ross in the Antarctic: the voyages ofJames Clark Ross in HerMajesty’s ships Erebus andTerror 1839-1843. Whitby Caedmon of Whitby, 276 pp.

SKULASON, S . & SMITH, T.B. 1995. Resource polymorpbisms in vertebrates. Trends in Ecology & Evolution, 10, 366-370.

SMITH, G.R. 1996. Introduction to classification. In BOND, C.E., Biology of fishes (2nd edition). Fort Worth: Saunders College Publishing, 59-75.

SMITH, G.R. &TODD, T.N. 1984. Evolution of species flocks of fishes in north temperate lakes. In ECHELLE, A.A. & KORNFIELD, I., eds Evolution of fish species flocks. Orono: University of Maine at Orono Press, 45-68.

SMITT, F.A. 1898. Poissons de I’expedition scientifique a la Terre de Fue. Bihang till Kongliga Svenska Vetenskaps-Akademiens Handlingar, 24(5), 1-80, plates I-VI.

STAM, W.T., BEINTEMA, J.J., D’AVINO, R., TAMBURRINI, M. & DI PRISCO, G. 1997. Molecular evolution of hemoglobins of Antarctic fishes (Notothenioidei). Journal of Molecular Evolution, 45, 437-445.

STAM, W.T., BEINTEMA, J.J., D’AVINO, R., TAMBURRINI, M., COCCA, E. & DI PRISCO, G. 1998. Evolutionary studies on teleost hemoglobin sequences. In DI PRISCO, G., PISANO, E. & CLARKE, A,, eds. Fishes of Antarctica: a biological overview, Milan: Springer Verlag, 355- 359.

VOSKOBOINIKOVA, 0,s. 1994. Rates of individual development of the bony skeleton of eleven species ofthe family Nototheniidae. Journal oflchthyology, 34, 108-120.

VOSKOBOINIKOVA, 0,s. 1997. Osteological development of the Channichthyidae (Teleostei: Notothenioidei). Cybiiim, 21,369-379.

WAITE, E.R. 1909. Vertebrata of the Subantarctic islands of New Zealand. In CHILTON, C., ed. The Subantarctic islands of New Zealand, vol. II. Wellington: Philosophical Institute of Canterbury, 542-600.

409-4 10.