Revision of the hirsuticornis‐like species of Macrothrix Baird, 1843 (Cladocera: Anomopoda: Macrothricidae) from Subantarctic and temperate regions of the southern hemisphere

The aim of this paper is to revise populations of Macrothrix cf. hirsuticornis (Cladocera: Anomopoda: Macrothricidae) from different regions of the southern hemisphere. It is demonstrated that M. hirsuticornis Norman and Brady, 1867 s. str. is absent there, and five related species occupy different Subantarctic islands and the southernmost portions of South America, and Africa. Macrothrix boergeni Studer, 1878 from the Kerguelen Archipelago is redescribed and a neotype is selected. All populations in the southernmost portion of continental South America, Tierra del Fuego, Falklands, South Georgia, South Orkney Islands, and on the Antarctic Peninsula belong to M. oviformis Ekman, 1900. All the taxa described from this region—M. ciliata Vávra, 1900, M. odontocephala Daday, 1902, M. propinqua Sars, 1909, and, probably, M. inflata Daday, 1902—are junior synonyms of M. oviformis. Two new species are established: M. sarsi sp. nov. from the Cape region of South Africa and M. ruehei sp. nov. from Crozet, Marion islands, and Île Amsterdam. Macrothrix cf. flagellata Smirnov and Timms, 1983, previously known only from Tasmania, is found on Macquarie Island too. Differences between species from the southern hemisphere and Palaearctic M. hirsuticornis are summarized. It is demonstrated that characters of the general body shape (i.e. presence of a hood or a tooth on posterior head border) have a limited value for the systematics of Macrothrix. In contrast, some fine details, mostly missed by previous authors, are valuable for species discrimination. The present study increases the number of species recorded from the Antarctic‐Subantarctic region. Probably, the current pattern of Macrothrix distribution results from a disruption of a pan‐continental (early Mesozoic?) species complex.


Introduction
A decade ago Korovchinsky (1996) pointed out the poor state of systematics for the Macrothricidae (Cladocera: Anomopoda). Fortunately the situation has significantly improved, especially for Macrothrix Baird, 1843, which was previously regarded as a ''hopeless'' genus (Smirnov 1976). Several species groups have been revised (Silva-Briano et al. 1999;Dumont et al. 2002;Kotov and Hollwedel 2004; and several remarkable new representatives have been found (Ciros-Pérez et al. 1996;Elías-Gutiérrez and Smirnov 2000). The dissertation of Silva-Briano (1998) was a significant contribution to the morphology and systematics of the genus, but, unfortunately, only some of its chapters have been published (Silva-Briano et al. 1999;Dumont et al. 2002). It is necessary to note that most of the aforementioned efforts were concentrated on tropical and subtropical regions.
Investigations on the Cladocera of the Subantarctic were started by Studer (1878), who described a series of species from the Kerguelen Archipelago including Macrothrix boergeni Studer, 1878. Subsequently, a series of new species and varieties was found in the southernmost portions of South America and Africa (Ekman 1900;Daday 1902;Sars 1916). Populations of Macrothrix sp. have also been reported from many Subantarctic localities (Pugh et al. 2002;Dartnall et al. 2005), and from the Antarctic Peninsula (Harding 1941).
While Smirnov (1976Smirnov ( , 1992 regarded all species of Macrothrix from the Subantarctic islands, southern areas of South America and Africa, as well as many other taxa from different regions of the planet, as junior synonyms of a widely distributed species M. hirsuticornis Norman and Brady, 1867. David G. Frey in a personal communication to Dartnall (1995) doubted the conspecifity of the Palaearctic and Subantarctic populations. Examination of populations from different regions of the southern hemisphere revealed substantial differences between them, sufficient to consider them to be a series of separate species, and also different from Palaearctic M. hirsuticornis s. str.

Material and methods
Samples of Macrothrix preserved in formalin or alcohol were obtained from different museums and colleagues. Animals were selected from preserved samples under a binocular stereoscopic microscope, placed on slides (in a drop of a glycerol-formaldehyde mixture) and studied under an optical microscope in toto. At least five parthenogenetic females from each locality (with the exception of the museum loans) were dissected under a stereoscopic microscope for the study of appendages and postabdomen. A single female was dissected from Sars's type material, with the permission of the Collection Manager. The dissected parts were kept on a series of slides, like a single paralectotype. Drawings were prepared using a camera lucida attached to an Alphaphot compound microscope.
A system of numeration for different setae on thoracic limbs proposed by Kotov (2000) for chydorids was used here, on the basis of similarity of limbs in Chydoridae and Macrothricidae (Smirnov 1971(Smirnov , 1976Dumont and Silva-Briano 1998). All operations with SEM have been described previously (Kotov 1999;Kotov and Š tifter 2006).
The cladistic analysis was performed using PAUP program version 4.0b10 for 32 bit Microsoft Windows (Swofford 2000), using branch-and-bond search with an aim to elucidate the possible phylogeny of the hirsuticornis-like species, with the well-studied M. tripectinata as an outgroup. A bootstrap simulation of 100 replications was performed as a test of robustness of this analysis.
The following abbreviations are used for the collections: AM, Australian Museum, Sydney, Australia; AAK, personal collection of A. A. Kotov Postabdomen subovoid, with rounded distal extremity, without a ''heel'' basally, and without a reticulation on sides. Ventral margin straight, with a few series of small, robust denticles. Dorsal margin distinctly bilobed; preanal margin with transversal series of minute   setules, anal margin with groups of thicker setules. Postabdominal seta with short distal segment, densely armed with relatively short setules; basal segment with very few (two to three) short, sparsely located setules. On external side of postabdominal claw, a series of three to five robust denticles; medial row (on ventral margin) of about seven to nine denticles; inner row with numerous denticles, organized in three successive series.
Antenna I widened distally, straight, without a subapical external angulation; sensory seta at distance of about two antennular diameters (at base) from antenna I joint; on anterior face about six to eight transverse rows of spinules, associated with distinct reticulation. Nine short aesthetascs, three of them slightly larger than the rest. Antenna II with distal burrowing spine on basal segment somewhat shorter than proximal segment of exopod. Length of all apical swimming setae subequal. Lateral seta on proximal endopod segment larger than other setae, lacking robust denticles in middle. A spine on second segment of exopod half as long as the next (second) segment. On posterior side of segments 1-3 of exopod there are a series of small additional denticles.
Limb I with outer distal lobe supplied with longest apical seta having distal segment unilaterally armed with relatively robust setules, located more rarely on proximal portion of seta distal segment; inner-distal lobe with three bisegmented setae of different size, unilaterally setulated in distal part, robustness of armature different in different setae, smallest one with numerous fine setules; two ejector hooks of different size, sometimes one of them rudimentary; a remainder of gnathobase I with single fully setulated setae. On limb II, scrapers 1-2 with delicate feathering, but scrapers 3-7 with denticles specially massive; a solitary posterior seta near gnathobase present; filter plate II with four setae, without a rudiment of fifth seta. On limb III epipodite with five setae; a distal group of three long setae, seta 1 shortest, armed with robust denticles; setulated projections proximally to seta 3 and between setae 2 and 3; on inner-distal portion, seta 1 with specially short and robust denticles; seta a with fine setules basally and robust spinules distally, seta b characteristically long; basal endite posteriorly with four soft setae. Limb IV with exopodite small, bearing only a distal group of three bilaterally feathered setae of subequal size; on inner-distal portion of this limb seta 1 with strong setules basally and two to three robust denticles distally; posteriorly a row of five long setae. On limb V there are three setae at inner margin, sometimes seta 3 reduced.
Size. Up to 1.06 mm.

Full redescription
Adult parthenogenetic female. Body of large female subovoid in lateral view ( Figures 1A, 2A, B, 6A), with maximum height in the middle, height/length50.66-0.70. Dorsal margin in general as a regular arch from tip of rostrum to posteriormost point, without traces of cervical depression or with a slight depression, mostly breached by a ''step'' in posterior boundary of head, sometimes forming a blunt tooth ( Figure 1A). Dorsal margin of valves not elevated, or slightly elevated above dorsal margin of head. Postero-dorsal angle as a rounded triangle, lies in level of middle of body height. Whole surface of valves with fine reticulation, while head without it. No remarkable structures on valves or head. In anterior view body moderately compressed laterally ( Figure 2C).
Head large, length/body length up to 0.4. In lateral view, dorsal margin in general evenly convex, no dome above eye; ventral margin almost straight, with minute crossing ridges, no projection at base of labrum ( Figure 1B, C). A special line (fold) goes from mandibular joint anteriorly, it corresponds to a poorly expressed fornix. Compound eye large, ocellus also large (its size more than half of eye diameter), located approximately in middle of distance between tip of rostrum and eye.
''Dorsal head pore'' (''dorsal organ'' or ''window'' are more correct names, because no pore is really present here, only a window of specialized cuticle) ovoid, small, located on posterior part of head ( Figure 2E). Labrum wide, in lateral view approximately triangular, with a moderately projected apex bearing several small tubercles, and setulated distal labral plate (term according to Kotov 1999, or anterior plug according to Dumont and Silva-Briano (2000)).
Valve surface with a distinct reticulation. Dorsal margin without any serration, but ventral margin denticulated. Marginal setae jointed to posterior sides of these denticles ( Figures 1D, G, 2D, 6B). These setae variable in length and size in different individuals, but there is a general order to their sequence: there are two ventrally directed setae between each larger, laterally oriented one, characteristic also of the majority of other species (Kotov 1999;Kotov and Hollwedel 2004;. In anterior and posterior portion of ventral margin the order of seta alternation is not too precise. Thorax long, while abdomen short, without projections. Intestine without convolutions. Postabdomen subovoid in lateral view, with a rounded distal extremity, without ''heel'' (inflated base of postabdominal setae) basally ( Figures 1H, 2H). Ventral margin straight, with few series of small, robust denticles. Dorsal margin distinctly bilobed, and the incision, which bilobes the margin, located at the level of proximal border of anus, separating anal and preanal margin. The latter long, slightly and regularly convex, with short transversal series of minute setules. On anal margin groups of setules with size significantly larger than those on postanal margin, laterally to them there are series of finer setules ( Figure 2G). Small postanal margin also with series of minute setules. A reticulation on sides of postabdomen fully absent.
Postabdominal seta approximately as long as postabdomen, with short distal segment, densely armed with relatively short setules; basal segment with a very few (two to three) short, sparsely located setules.
Postabdominal claw small, directed distally, slightly and regularly bent dorsally, with pointed tip and wide base in lateral view. On claw, a series of few (three to five) robust denticles on external side, medial row (immediately on ventral margin as seen laterally) of about seven to nine denticles (Figures 1I,J,3A,B), and inner dorsal row with numerous denticles, organized in three successive series ( Figure 1K).
Antenna I widened distally, straight, without subapical external angulation (''subapical ventral angulation'' sensu of Silva-Briano (1998)); sensory seta located externally at distance of about two antennular diameters (at base) from antenna I joint (Figures 1L,M,3C). About six to eight transverse rows of spinules, associated with a distinct reticulation on anterior surface of antenna I, series of fine spinules at its end. Nine relatively short terminal aesthetascs (longest about one-quarter of antenna I length), each with two minute ''claws'' at the apex ( Figure 1N, O). Three aesthetascs slightly larger than the rest, bearing unknown thin-walled, delicate structures, which could be additional sensory elements, but there is a chance that these are only epibiotic bacteria.
Antenna II large, coxal region folded, with two small basal sensory setae of slightly different size in middle part and rows of small setules on each fold ( Figures 3D, 4A, 6C). Basal segment robust, bearing numerous transverse series of spinules, and long, bisegmented distal sensory seta at inner (posterior) margin. Distal burrowing spine somewhat shorter than proximal segment of exopod, located on outer (anterior) surface, close to end of the basal segment ( Figure 3E).
Antennal branches long (about two times longer than basal segment), only proximal member of exopod shortened, all other segments elongated. Segments 2-3 of exopod subequal in size, while terminal segment as long as these two. Endopod apical segment longer than each of the other segments of this branch. All segments with transverse rows of setules. Swimming setae 0-0-1-3/1-1-3, spines 0-1-0-1/0-0-1. Length of all apical swimming setae subequal, approximately equal to length of basal segment plus length of branch. Each seta is marked with an individual number in Figure 4A, armature of each seta is illustrated in Figure 4B-I. Lateral seta on proximal endopod segment larger than other setae, lacking robust denticles in middle (Figures 3F,. Apical spines short, from slightly curved to straight. A single spine on second segment of exopod, this spine half as long as next segment. On posterior side of exopod segments 1-3 there are a short series of small additional denticles (see discussion in ).
Mandible small, elongated, evenly dilated distally. Mandibular articulation located externally at point where valve and head come together ( Figure 2F).
Limb I large, without accessory seta; outer distal lobe cylindrical ( Figures 3H, 4N, O), supplied with a long apical seta with distal segment unilaterally armed with relatively robust setules, located more rarely in proximal portion of seta distal segment, and a small lateral seta with bilaterally setulated distal segment. Inner-distal lobe massive, with median series of setules, few groups of minute setules, and three bisegmented setae of different size, unilaterally setulated in distal part, robustness of armature different in different setae, smallest one bears numerous fine setules. Endite III with a long, slightly curved, distally setulated seta a, while setae b-c short, clearly bisegmented, setulated bilaterally in middle ( Figures 3G, 4N, P); anteriorly on this endite there is a short bisegmented seta 1, bearing sparse, long setules distally ( Figure 4Q). Endite II with three long bisegmented setae of subequal size ( Figure 4P: d-f), each with distal segment bearing fine setules on distal parts, and more robust setules basally, and with fully setulated basal segment, and a fork-like seta 2 ( Figure 4R) anteriorly. Endite I with two bisegmented setae ( Figure 4P: g-h) naked basally and supplied with long, dense setules distally; a fork-like seta 3 ( Figure 4S) anteriorly. Ejector hooks of different size, bilaterally setulated, in some specimens second hook significantly reduced in size and naked ( Figures 4T-V). A fully setulated seta at inner side of limb base, so-called maxillar process, represents a remnant of gnathobase I (Kotov 1999).
Limb III: epipodite very small, globular, exopodite large and flat, with a distal group of three long setae ( Figure 5D: 1-3), seta 1 shortest and armed with robust denticles, setae 2-3 with fine setules; there are setulated projections proximally to seta 3 and between setae 2 and 3; lateral group consists of two setae (4 and 5), similar in armature to 2 and 3. Distal endite (see discussion of its homology in Kotov (1999)) anteriorly with three bisegmented setae ( Figure 5E: 1-3), unilaterally armed in distal part, seta 1 with specially short and robust denticles; small sensillum near each base of seta 1 and 3. Posteriorly, there are three soft setae: seta a with fine setules basally and robust spinules distally, setae b and c with fine setules distally, seta b characteristically long. Basal endite approximately equal in size to distal one. Anteriorly, a bottle-shaped sensillum and four setae with size increasing basally (4-7), each fully setulated distally, with distal segment longer than basal one. Posteriorly, four thick, soft setae (d-g) subequal in size, each armed with fine setules, each has an inflated basal portion and a blunt tip. Gnathobase unclearly demarcated from basal endite, with four elements of distal armature: a large, bottle-shaped sensillum near border with basal endite, and two hooks plus a short bisegmented seta distally ( Figure 5E: 1-4).
Limb IV: pre-epipodite small, with few setules, epipodite small and globular, exopodite small, with only a distal group of three bilaterally feathered setae of subequal size ( Figure 5F: 1-3). Inner margin of limb with four setae ( Figure 5G: 1-4), seta 1 with strong setules basally and two to three robust denticles distally ( Figure 5H), while setae 2-4 each with an inflated basal and elongated distal part, the latter pointed at the tip, fully feathered. A small sensillum near each base of seta 2 and 3. Posteriorly, a row of five long, erect, soft setae, similar in size, bilaterally setulated from base to tip ( Figure 5G: a-e). Distal armature of gnathobase with four elements: a thick, bottleshaped sensillum near border with basal endite; a large seta with inflated basal segment and elongated, fully setulated distal segment; a heavy hook; and a small, naked, bisegmented and curved seta bearing minute setules distally. Posteriorly on gnathobase, a single small seta continues the posterior row of setae of the inner limb face, the sole vestige of a filter plate IV.
Limb V: pre-epipodite relatively small, flat, its margin setulated; epipodite large, subglobular. Only a small lobe with single seta represents a vestige of the exopodite. Innerdistal portion as a large flap, fringed by fine setules, on inner margin three setae with size significantly increasing distally ( Figure 5J: 1-3), in some specimens seta 3 almost reduced ( Figure 5I).
Differences of juvenile female. In contrast to adult, body somewhat lower (height/ length50.63-0.67), subquadrangular, without a tooth on posterior border of head, with valve dorsal margin straight, significantly elevating beneath head ( Figure 1P), with posteroventral angle oblique, located above middle axis of body, with antennae II and swimming antennal setae longer, rows of setules on antenna I and II weakly developed.
Ephippial female, male. Unknown. Gay (1981) attempted to described some peculiarities of the adult male, but no valuable information was presented.
Taxonomic comment. The only character of M. boergeni important for the differentiation from some other hirsuticornis-like species, reported by Studer (1878), is the finely setulated distal segment of the seta on the proximal segment of antenna II. During the whole of the 20th century (Rü he 1914;Smirnov 1976Smirnov , 1992, M. boergeni was regarded as a junior synonym of M. hirsuticornis; the former is shown to be a well-differentiated valid species (see Table I), the second hirsuticornis-like taxon described for the World's fauna.

Type material
Most probably lost. Absent from Ekman's Collection in SMNH (Kotov and Gololobova 2005). Macrothrix inflata Daday, 1902-syntypes: four females from unknown locality in Santa Cruz, Argentina, coll. by F. Sylvestri, slide DAD II/P-418. We found that Smirnov (1992) was correct when he said that the slide is dried, and the specimens are in a bad state.
Antenna I widened distally, slightly curved, without a subapical external angulation; sensory seta at distance of about 2-2.5 antennular diameters (at base) from antenna I joint; on anterior face about five to six transverse rows of strong spinules, and rows of finer setules, but no reticulation (Figures 7B, 8B, G, 9N, O, 13B). Nine short aesthetascs (longest shorter than half of antenna I length), two of them significantly larger than the rest. On the latter, sometimes there are some thin-walled, delicate structures ( Figure 9O), which could be additional sensory elements, but there is a chance that these are only epibiotic bacteria, as in the case of M. boergeni. Antenna II with distal burrowing spine on basal segment longer than proximal segment of exopod ( Figures 10A, 13D). Length of all apical swimming setae subequal, their armature represented in Figure 10B-I. Lateral seta on proximal endopod segment larger than other setae, with robust denticles in the middle (Figures 7F-G, I, N, P, 8I-K, 10J-N). A spine on second segment of exopod approximately half as long as next segment. On posterior side of segments 1-3 of exopod there are minute additional denticles.
Limb I with longest apical seta of outer distal lobe ( Figure 11A, B) having distal segment unilaterally armed with relatively fine, spaced setules; inner-distal lobe with three bisegmented setae of different size, unilaterally setulated in distal part, robustness of armature different in different setae, smallest one with whole distal segment setulated ( Figure 11C), or these setules present only in proximal portion of distal segment ( Figure 11B); two ejector hooks of similar size; a remainder of gnathobase I with single fully setulated setae. On limb II ( Figure 11D, E), scrapers 1-2 with delicate feathering, scrapers 3-7 with robust denticles of size characteristic for the genus; a solitary posterior seta near gnathobase present; filter plate II with four setae, without a rudiment of fifth seta. On limb III ( Figure 11F) epipodite small, exopod with a distal group of three long setae, seta 1 shortest, armed with short setules; setulated projections proximally to seta 3 and between setae 2 and 3; inner-distal limb portion ( Figure 11G) with seta 1 with short and robust denticles; seta a with fine setules basally and robust spinules distally, seta b longer than c; basal endite posteriorly with four soft setae. Limb IV ( Figure 11H) with exopodite small, bearing only a distal group of three bilaterally feathered setae of subequal size; on inner-distal portion of this limb ( Figure 11I) seta 1 with strong setules basally and two to five robust denticles distally; posteriorly row of five long setae. On limb V there are three setae on inner margin ( Figure 11J). Ephippial female. In contrast to adult parthenogenetic female, body less high (height/ length50.65-0.72), postero-dorsal angle more prominent and lies in level of middle of body height ( Figure 12A). Body less compressed laterally and lacking a dorsal keel ( Figure 12B). Dorsal portion of valves forming an ephippium, slightly pigmented in brownish colour, without clear border between it and rest of valves. Dorsal wall of carapace forms a special dark, chitinized plate. Two eggs in ephippium.
Male. Body subquadrangular in lateral view. Apex of labrum with a series of tubercles; a reticulation on sides of postabdomen fully absent, gonopores open on its distal margin, distalmost denticle in median series on postabdomen claw significantly larger than the rest. Antennular sensory seta located there at distance of about 1.5 antennular diameters (at base) from antenna I joint; at the same level, a large male seta, with length of half of antenna I length, bisegmented, with basal segment short and naked, and distal segment fully setulated; 14 relatively short terminal aesthetascs, two aesthetascs slightly larger than the rest. Limb I with inner-distal lobe bearing four setae, additional male seta large; copulatory hook large, with three ridges distally, copulatory brush seta on a low pedestal, bisegmented, with blunt tip to distal segment.
Size. Females up to 1.14 mm, males up to 0.56 mm.

Redescription of adult male
Body subquadrangular in lateral view, height/length50.63-0.69, dorsal margin without a distinct cervical depression, no ''step'' in posterior boundary of head ( Figure 12C). Dorsal margin of valves not elevated significantly above dorsal margin of head. Postero-dorsal angle obtuse, rounded, lies dorsal to middle of body height.
Head with dorsal margin evenly convex, no dome above eye; ventral margin slightly concave to straight, without ridges, no projection at base of labrum ( Figure 12D). Compound eye large, ocellus even smaller than in female (its size about one-third of eye diameter), located closely to tip of rostrum. Dorsal organ small, located on posterior part of head.
Labrum in lateral view approximately triangular, with a moderately projected apex bearing several small tubercles, and setulated distal labral plate.
Valve surface with a delicate reticulation. Dorsal margin without any serration, ventral margin as in female, but marginal setae relatively longer ( Figure 12E, F).
Postabdomen in lateral view ( Figure 12G) with subquadrangular distal portion and without a ''heel'' basally. Ventral margin inflated, without series of denticles. Dorsal margin distinctly bilobed, preanal margin regularly convex, with short transversal series of minute setules. On anal margin groups of setules with size significantly larger than those on postanal margin, laterally to them there are series of finer setules. A reticulation on sides of postabdomen fully absent. Gonopores open on distal margin, ventrally to postabdominal claws.
Postabdominal seta longer than postabdomen, with short distal segment, densely armed with long setules; basal segment with numerous, shorter setules.
Postabdominal claw small, directed distally, less elongated than in female, slightly and regularly bent dorsally, with pointed tip and wide base in lateral view ( Figure 12H). On claw, a series of seven to ten fine denticles on external side; medial row with about five to six denticles, distalmost one significantly larger than the rest.
Antenna I almost straight, with maximum width in basal portion, this part a homologue of subapical external angulation; antennular sensory seta located at distance of about 1.5 antennular diameters (at base) from antenna I joint ( Figure 12IJ). On anterior surface of antenna I, at level of sensory seta, a large male seta, half length of antenna I, bisegmented, with basal segment short and naked, and distal segment fully setulated. About six to seven transverse rows of spinules on anterior surface of antenna I, but no reticulation. Fourteen relatively short terminal aesthetascs (longest one-quarter length of antenna or even shorter), apex of each with a conical depression on distal side ( Figure 12K, L), two aesthetascs slightly larger than the rest.
Antenna II with relative size larger than in female; basal segment of exopod relatively shorter than in female; distal burrowing spine on basal segment and a spine on the second segment of exopod significantly longer than in female ( Figure 10O).
Limb I with outer distal lobe as in female; inner-distal lobe with four setae, additional male seta large; copulatory hook large, with three ridges distally ( Figure 12M). Copulatory brush seta on a low pedestal, bisegmented, with blunt tip to distal segment.
Juvenile (pre-reproductive) male. Body shape ( Figure 7Q) similar to adult; height as in adult male (height/length50.61-0.66), postabdomen ( Figure 7R) as in female, distal denticle of median row on postabdominal claw ( Figure 7S) not as large as in adult male; antenna I with nine aesthetascs and shorter male seta ( Figure 7T, U); male seta on inner-distal lobe of limb I smaller and located at a distance from other setae, copulatory hook shorter than in adult ( Figure 7V).  Sars (1909), female length reaches 1.14 mm.
Taxonomic notes. The southernmost portion of the South American continent, South Atlantic (Subantarctic) islands and Antarctic Peninsula are inhabited by only a single hirsuticornis-like species, for which several formal names were suggested: M. oviformis Ekman, 1900, M. ciliata Vávra, 1900, M. inflata Daday, 1902, M. odontocephala Daday, 1902, and M. propinqua Sars, 1909 Two species (M. oviformis Ekman and M. ciliata Vávra) were published within the same year; the article of Ekman (1900) appeared on 22 October 1900 (this information present on the title page of the issue where it was published). Attempts to establish when Vávra's (1900) paper appeared, including special requests to several libraries in Germany (including Hamburg, where it was published), were unsuccessful. According to case 21.3.2 of the International Code of Zoological Nomenclature (International Commission on Zoological Nomenclature 2000), the latter article must be regarded as published on ''the last day of the year'', namely on 31 December 1900. So, M. oviformis Ekman, 1900 has priority, and M. ciliata is its junior synonym.
First description of M. oviformis Ekman, 1900 was relatively detailed for that time. The distinctive characters listed by Ekman (1900, p 71-73) are now considered of dubious value. The first description contains some important characters such as (1) a depression bordering dorsal margin of head and valves; (2) nine aesthetascs of different size; (3) two dorsal projections on abdomen, and his illustrations adequately represent others such as (1) shallow cervical depression; (2) presence of numerous setules on proximal segment of postabdominal seta; (3) long setules on distal segment of postabdominal seta. Specimens very similar to that illustrated by Ekman (1900) were found at localities close to the type locality (Rio Turbio), as well as at other localities in continental South America.
Macrothrix ciliata Vávra, 1900 was described from the Falkland Islands. Most of the characters listed by Vávra (1900, p 18-19) are characters of a juvenile hirsuticornis-like species. A neotype of this taxon is here selected from Ekman's (1905) sample collected in the vicinity of the type locality; a juvenile female is selected, confirming Vávra's ideas on this species. There is no doubt that Macrothrix ciliata is a junior synonym of M. oviformis.
Macrothrix inflata Daday, 1902 is more problematic. Females from a single population from continental South America (DGF 8780, unfortunately, with few specimens), lacking a depression between head and valves, also have some other specific traits ( Figure 7A-G): (1) setules on seta on proximal segment of antennal endopod are relatively small, and (2) the dorsal organ is very large as compared with populations with clear border between head and valves. This may indicate that M. inflata is a valid species, but this conclusion must be confirmed by examination of better material; however, there is a high probability that the inflata-like morphotype is only an extreme example of the oviformis variability. Ekman (1900) clearly stated that M. oviformis has a distinct border between head and valves, so in the case that two taxa are recognized in South America, M. oviformis is widely distributed and M. inflata rare.
Macrothrix propinqua Sars, 1909 was described in detail, but no real differences from other species described from the South Atlantic region were reported. Re-examination of Sars' (1909) type material led to the conclusion that his taxon is an apparent synonym of M. oviformis.
Two other species described from the southern part of South America, M. goeldi Richard, 1897 and M. magna Daday, 1902, are not members of the hirsuticornis-group. Macrothrix goeldi Richard, 1897 is a member of the laticornis-or spinosa-group, because the first description (Richard 1897, p 287-289) contains a clear reference to distinct denticulation at the dorsal margin of the valves. Macrothrix magna Daday, 1902 was regarded by Smirnov (1992) as a junior synonym of M. hirsuticornis, but this is incorrect. The male of M. magna has a very thin antenna I with a great distance between male seta and antennular sensory seta. This is apparently a relative of the triserialis-rosea group, sensu Dumont et al. (2002). At the same time, M. magna has postabdominal seta with a long distal segment, so the former is not a junior synonym of M. triserialis s. str. or M. elegans Sars, 1901 (the latter is one of the most common species in the more southern portion of South America; . Macrothrix magna needs to be revised, as a species outside of the hirsuticornis-like group. Olivier (1962) listed differences between all species described by previous authors, but these differences are dubious.
Distribution. Southernmost portion of continental South America, Tierra del Fuego, Falkland Islands, South Georgia, South Orkney Islands, and the Antarctic Peninsula. All records from the South Atlantic listed by Pugh et al. (2002) refer to this species. It is necessary to note that in these localities M. oviformis is a quite common species. Most probably Macrothrix from the South Shetland Islands (Toro et al. 2007) and Palmer Land (Pugh et al. 2002) also belong to this species.

Etymology
The species is named after George Ossian Sars, a famous Norwegian carcinologist and one of the greatest investigators of the Cladocera, whose material is used for the species description.

Type locality
A water body ''Cape of Good Hope V'', Western Cape Province, Republic of South Africa. Sars (1916) wrote that the sample was taken ''in the neighbourhood of Bergvliet''.

Diagnosis
Adult parthenogenetic female. In lateral view body subovoid, cervical depression very shallow, dorsal margin not breached by a ''step'' in posterior boundary of head ( Figure 14A), dorsal margin of valves with a minute serration ( Figure 14D), posterodorsal angle as rounded triangle, lies somewhat ventrally to middle of body height. In anterior view, body compressed laterally and supplied with a sharp dorsal keel. No dome above eye. Ocellus small. Dorsal organ small. Labrum with a slightly projected apex bearing two tubercles. Setae on ventral margin of valve as in other species (Figures 14E-G).
Postabdomen subovoid, with rounded distal extremity, without ''heel'' basally, distinct reticulation on sides ( Figure 14H), its ventral margin straight, with few series of fine spinules; first series consists of specially long elements ( Figure 14I, J). Dorsal margin distinctly bilobed; preanal margin with transversal series of minute setules, anal margin with groups of thicker setules. Postabdominal seta longer than postabdomen, with distal segment somewhat shorter than basal one and densely armed with long setules; basal segment with numerous, short setules. On external side of postabdominal claw, a series of three to four thin denticles.
Antenna I markedly widened distally, slightly curved, without a subapical external angulation; sensory seta located externally very far (three or even more antennular diameters at base) from antenna I joint; on anterior face five transverse rows of spinules, but no reticulation ( Figure 14K, L). Nine relatively long aesthetascs, two of them significantly larger than the rest. Antenna II with distal burrowing spine on basal segment markedly longer than proximal segment of exopod ( Figure 15A). Length of all apical swimming setae subequal, armature as in other species ( Figure 15D-G). Lateral seta on proximal endopod segment larger than other setae, lacking robust denticles in middle ( Figure 15B, C). A spine on second segment of exopod short, less than half length of next segment. On posterior side of segments 1-3 of exopod there are no additional denticles, but sometimes there are several small spinules.
Limb I with apical seta of outer distal lobe having distal segment unilaterally armed with robust setules ( Figure 16A, B), inner-distal lobe with three bisegmented setae of different size, smallest seta setulated only on its middle portion, anterior setae as represented in Figure 16CE, posterior setae d-e with bulbs at tips ( Figure 16F), ejector hooks of subequal size ( Figure 16G, H). On limb II, scrapers 1-2 with delicate feathering, scrapers 3-7 with thicker denticles of size characteristic for the genus; a solitary posterior seta present near gnathobase ( Figure 16J, K), filter plate II with four setae, without a small rudiment of fifth seta. On limb III epipodite with five setae; seta 1 as long as 3, armed with fine setules distally; smooth setulated projections proximally to seta 3 and between setae 2 and 3 ( Figure 16L), on inner-distal limb portion, seta 1 with fine setules distally, seta a with fine setules basally and strong spinules distally, seta b slightly longer than a, but markedly longer than c ( Figure 16M), basal endite normally with four soft setae ( Figure 16L). Limb IV with exopodite bearing a distal group of three bilaterally feathered setae of somewhat different size ( Figure 16N); on inner-distal portion of the limb seta 1 with strong setules basally and five to six denticles distally; posteriorly on limb inner margin a row of five long setae ( Figure 16O). On limb V there are three setae at inner margin, seta 1 with distalmost extremity as a naked whip ( Figure 16P).
Atypical adult. A single atypical adult had two setae on exopodite II ( Figure 16J) and five setae on basal endite III ( Figure 16M).
Juvenile female. In contrast to adult, body more elongated and quadrangular ( Figure 14M), with posterior border of dorsal head margin elevated above valves; posterior margin of valve almost straight, postero-dorsal angle at level of dorsal margin, ventral margin of valve strongly convex; antenna I relatively longer; antennae II and swimming antennal setae longer, rows of setules on antenna I and II weakly developed.
Taxonomic notes. Sars (1916) determined this population from the Cape of Good Hope as belonging to his species M. propinqua Sars, 1909. The latter is here considered as a junior synonym of M. oviformis. At the same time, the South African populations, misidentified as ''M. propinqua'' by Sars (1916), belong to another species, named here M. sarsi sp. nov.
The description of Sars (1916) contains few characters helpful for distinguishing the species of hirsuticornis-like forms as some of the ''differences'' are doubtful in terms of recent understanding of Macrothrix systematics. For example, Sars (1916) mentioned a large epipodite on limb V, but its size is too variable in other species of Macrothrix  to be a good taxonomic character. The present opinion is based predominantly on a study of Sars' material instead of his description.
Distribution. Macrothrix sarsi sp. nov. is present only in the Cape of Good Hope region of the Republic of South Africa. Some species of Macrothrix (i.e. M. spinosa, M. odiosa, and M. capensis) are common in South African water bodies. In contrast, M. sarsi is apparently a rare species, it was not found by Smirnov (2007a), who examined in detail some 290 samples from South Africa. Clarke and Rayner's (1999) report of ''M. propinqua'' from Namibia must be confirmed.

Etymology
The species is named after Friedrich Eduard Rü he, well-known German investigator of Cladocera, who identified this species as M. hirsuticornis from Île Amsterdam (Rü he 1914).

Type locality
A pond at Baie Americaine, Île de la Possession, Crozet Islands, French Subantarctic Territories. The type series was collected 24 February 1968 by L. Davies.

Diagnosis (based exclusively on Crozet specimens)
Parthenogenetic female. In lateral view body subovoid (height/length50.62-0.67 in large adults), cervical depression present, dorsal margin breached by a ''step'' in posterior boundary of head, forming a blunt tooth, dorsal margin of valves not elevated under dorsal margin of head, without any serration ( Figure 17A). Postero-dorsal angle as rounded triangle, lies in level of middle of body height. Body compressed laterally, with a low, but sharp dorsal keel ( Figure 17B). No dome above eye; ocellus small ( Figure 17D). Dorsal organ ovoid, small. Labrum with a thick, rounded, moderately projected apex lacking any tubercles ( Figure 17D, E). Armature of ventral margin as in other species ( Figure 17F, G).
Postabdomen without ''heel'' basally, and without a reticulation on sides. Ventral margin straight, with few series of small denticles. Postabdominal seta with a short distal segment, armed with relatively short setules; basal segment without any setules ( Figure 17L, M). On external side of postabdominal claw, a series of six to nine moderately robust denticles; medial row of about seven to nine denticles; inner row with numerous denticles, organized in two successive series ( Figure 17H, I), on inner side two pectens of denticles ( Figure 17K).
Antenna I widened distally, straight or slightly curved, without a subapical external angulation; sensory seta at distance of about 1.5-2 antennular diameters (at base) from antenna I joint; on anterior face about five to seven transverse rows of spinules, but no reticulation ( Figure 17N, O). Nine short aesthetascs, two or three of them significantly larger than the rest. Antenna II with distal burrowing spine on basal segment as long as or longer than proximal segment of exopod ( Figure 18A). Length of all apical swimming setae subequal, their armature as in M. boergeni. Lateral seta on proximal endopod segment larger than other setae, lacking robust denticles in middle ( Figure 18G). A spine on second segment of exopod half as long as next segment ( Figure 18B, C), on posterior side of segments 1-3 of exopod there are series of relatively large additional denticles.
Limb I outer distal lobe with longest apical seta having distal segment unilaterally armed with robust setules; inner-distal lobe with three bisegmented setae of different size, unilaterally setulated in distal part, smallest one with whole distal segment setulated, or these setules present only in proximal portion of distal segment ( Figure 18H); two ejector hooks of similar size ( Figure 18I). On limb II, scrapers 1-2 with delicate feathering, scrapers 3-7 with robust denticles of size characteristic for the genus ( Figure 18J, K); a solitary posterior seta present near gnathobase; filter plate II with four setae, without a rudiment of fifth seta. On limb III epipodite with five setae; a distal group of three long setae, seta 1 armed with robust denticles ( Figure 18L); setulated projections proximally to seta 3 and between setae 2 and 3; on inner-distal limb portion, seta 1 with short and robust denticles; seta a with fine setules basally and robust spinules distally, seta b only somewhat longer than c ( Figure 18N); basal endite posteriorly with four soft setae. Limb IV with exopodite small, bearing only a distal group of three bilaterally feathered setae of different size ( Figure 18O); on inner-distal portion of this limb seta 1 with strong setules basally and two to six robust denticles distally ( Figure 18P, Q); posteriorly, row of five long setae. On limb V there are three setae at inner margin.
Differences of juvenile female. In contrast to adult, body more elongated, with less-developed tooth on posterior border of head, with valve dorsal margin less convex, with antennae II and swimming antennal setae longer, rows of setules on antenna I and II weakly developed ( Figure 17C).
Ephippial female, male. Unknown. (3) shorter aesthetascs of antenna I with smaller differences between smallest and largest members; (4) very small additional denticles on segments 1-3 of antenna II; (5) more prominent projections between setae 2 and 3 and distally to seta 3 on exopodite III; (6) more numerous denticles (up to six) on seta 1 of inner margin of limb III. Except for difference 1, the rest of the specific traits of Île Amsterdam populations seem to be subject to strong variability among Macrothrix populations. Populations from all three groups of islands examined here are considered conspecific.
Taxonomic notes. Rü he (1914) was the first to observe M. ruehei sp. nov., which he identified as ''M. hirsuticornis'', although he mixed in the description populations from Île Amsterdam and Kerguelen, belonging to different species in present understanding. Brehm (1958) made a drawing of the head of ''M. hirsuricornis'' from Île Amsterdam with widely rounded labral apex. This is undoubtedly M. ruehei sp. nov. There is no doubt about the conspecificity of populations from Crozet, Marion, and Île Amsterdam, but due to small differences between them, all material not from Crozet is ''expressly excluded from the type series'' (case 72.4.1 of the International Code of Zoological Nomenclature (International Commission on Zoological Nomenclature 2000)).
Distribution. French Southern Territories: Crozet, Marion, Île Amsterdam, but not Kerguelen, where its congener, M. boergeni, occurs. Apparently, M. ruehei sp. nov. is a very common species in small water bodies on the aforementioned islands. Note that it is not clear which species of Macrothrix (recorded by Dartnall (1995)) inhabits Heard Island.
Diagnosis (based exclusively on population from Macquarie) Parthenogenetic female. In lateral view body subovoid, relatively high (height/length50.65-0.70), cervical depression absent, dorsal margin not interrupted by a ''step'' in posterior boundary of head, dorsal margin of valves without any serration, postero-dorsal as rounded angle, lies in level of middle of body height ( Figure 19A). No dome above eye. Ocellus small, less than half of eye diameter. Dorsal organ ovoid, small. Labrum with a strongly projected triangular apex bearing few tubercles ( Figure 19B). Valves with fine reticulation ( Figure 19C) and marginal setae as in previously described congeners ( Figure 19D).
Postabdomen subovoid, with rounded distal extremity, without ''heel'' basally, with a reticulation on sides ( Figure 19E). Ventral margin straight, with four to five series of moderately long, thin spinules. Dorsal margin distinctly bilobed; preanal margin with transversal series of minute setules, anal margin with groups of thicker setules. Postabdominal seta distal segment slightly shorter than basal one, densely armed with relatively short setules; basal segment with numerous, short setules. On external side of postabdominal claw, a series of 10-12 fine spinules; medial row of about six to eight denticles ( Figure 19F).
Antenna I very slightly widened distally, straight to slightly curved, without any traces of subapical external angulation; sensory seta at distance of about 1.5-2 antennular diameters (at base) from antenna I joint; on anterior face about seven to ten transverse rows of spinules, but no reticulation ( Figure 19G). Nine short aesthetascs, three of them significantly larger than the rest ( Figure 19H). Antenna II with distal burrowing spine on basal segment slightly shorter than proximal segment of exopod ( Figure 20A), distal sensory seta ( Figure 20B) somewhat longer than distal burrowing spine. Length of all apical swimming setae subequal. Lateral seta on proximal endopod segment larger than other setae, lacking robust denticles in middle ( Figure 20C-G). A spine on second segment of exopod longer than half of next segment. On posterior side of segments 1-3 of exopod there are a series of small, thin additional denticles.
Limb I outer distal lobe with longest apical seta having distal segment unilaterally armed with robust setules, inner-distal lobe with three bisegmented setae of different size, unilaterally setulated in distal part, smallest one with whole distal segment and distal portion of basal segment setulated; two ejector hooks of different size; maxillar process with two setae (Figure 20H), anterior setae as represented in Figure 20I-K. On limb II, scrapers 1-2 with delicate feathering, scrapers 3-7 with robust denticles of size characteristic for the genus; a solitary posterior seta near gnathobase; filter plate II with four setae, without a small hillock in position of rudiment of fifth seta ( Figure 20L). On limb III epipodite with a distal group of three long setae, seta 1 shortest, armed with robust denticles; setulated projections proximally to seta 3 and between setae 2 and 3 ( Figure 20M); on inner-distal limb portion, seta 1 with short and robust denticles; seta a with fine setules basally and robust spinules distally, seta b markedly longer than c; basal endite posteriorly with four soft setae. Limb IV ( Figure 20N, O) with exopodite small, bearing only a distal group of three bilaterally feathered setae of different size; on inner-distal portion of this limb seta 1 without setules basally and few denticles distally; posteriorly row of five long soft setae. On limb V there are three setae at inner margin ( Figure 20P).
Size. In our samples 0.62-1.22 mm. In Tasmania, M. flagellata specimens up to 1.39 mm are found (Smirnov and Timms 1983). Taxonomic notes. At this level of morphological description of M. flagellata, it is impossible to conclude that I found just M. flagellata s. str. An accurate comparison with the original type material should be made and the Australian macrothricids need to be reassessed. Smirnov and Timms (1983) initially placed M. flagellata in the genus Echinisca because of the undilated antenna I; however, this genus was subsequently synonymized with Macrothrix (Smirnov 1992). Serious doubts were expressed regarding the subjective status of the character ''dilated-undilated antenna I'' (Kotov 1999). This character is applicable only in extreme states. In the intermediate state it is not possible to conclude whether antenna I of this species (or even specimen) is dilated or not. In the case of M. flagellata, a completely undilated antenna I is the only character distinguishing this species from others listed above. At the same time, M. flagellata and other aforementioned species have a set of similar characters, possible synapomorphies. In other hirsuticornis-like species the level of dilation is different, and variable. Consequently, I believe that M. flagellata is a hirsuticornislike species.

Type material
Lectotype: a female in bad condition from ''Ashborne, Sunderland'', Norman's Collection, NHM 1911.11.8  Ephippial female. Similar to parthenogenetic female, with slightly pigmented ephippium bearing a slight additional sculpture of fine polygonal reticulation, no clear border between ephippium and rest of valve, dorsal wall of carapace forms a special dark, chitinized plate; normally two eggs, sometimes three in ephippium (Berg 1933;Hudec 1983).
Taxonomic notes. Initial examination of a relatively limited set of samples with Macrothrix from the Palaearctic (A. A. Kotov, unpublished) revealed three hirsuticornis-like taxa: M. hirsuticornis Norman and Brady, 1867, M. tripectinata Weisig, 1934, and M. dadayi Behning, 1941. Silva-Briano (1998 redescribed M. hirsuticornis s. str. in his PhD thesis, but these results are not published. It is necessary to note that although Silva-Briano (1998) concluded that M. cornuta is a ''horned'' morphotype of M. hirsuticornis, the female of the former species from Turkey (whose identity with Daday's (1903) species from Mongolia must be checked!) has strong spines on the seta on the proximal segment of the endopod of antenna II. Most probably the M. cornuta of Silva-Briano (1998) is another Palaearctic hirsuticornis-like species, but it is clear that the revision of M. hirsuticornis in the Palaearctic still has some way to go and must be done using better material.
Previous efforts by taxonomists examining the M. hirsuticornis group have predominantly concentrated on the variability of characters regarded as helpful for systematics by cladocerologists of the 19th century (Sars 1890;Lilljeborg 1901). These include the expression of a tooth or hood on posterior border of head, or additional spines on antennal branches (Fox 1962;Flö ssner 1967Flö ssner , 1972Hudec 1983;Margaritora and Usai 1983;Silva-Briano 1998;Flö ssner 2000). All of these publications, dealing with external shape of body, lacked a detailed analysis of other, ''finer'', traits of different populations. Only after such accurate examination of many populations may a conclusion be made on the number of hirsuticornis-like species in the Palaearctic.
In any case, all southern hemisphere species described above differ from Palaearctic M. hirsuticornis s. str. (Table I).
Distribution. Widely distributed in the Palaearctic. Status of Nearctic and all southern hemisphere populations must be checked.

Phylogeny
A list of characters analysed is given in Table I. Autapomorphies, marked by bold type, were excluded from the analysis; each variable character is marked as data missing. A cladistic search using 12 characters reveals three equally short trees (TL516; CI50.75; RI50.71); a strict consensus tree is represented in Figure 23. The 50% majority rule bootstrap simulation led to a tree of similar topology with the consensus tree. Due to this fact, branch probabilities were assigned to the aforementioned consensus tree.
Two well-supported clades are found: boergeni-ruehei and sarsi-oviformis-hirsuticornis; these are grouped together, but this grouping has a moderate support (66%). Macrothrix flagellata is not grouped with any other taxa; this species as well as M. tripectinata (the outgroup) are basal members of the studied group.

Morphology and systematics
Characters such as general body shape (the presence of a hood or tooth on the posterior head border) have a limited application for Macrothrix systematics (Fox 1962;Silva-Briano 1998). The aforementioned tooth is, probably, an element of antipredator morphology (Kotov 1999) in addition to spines on the branches of antenna II. Also, the general body shape in this genus is the subject of strong instar variability (Smirnov 1992). In contrast, some fine details, often neglected by previous investigators, are valuable for species determination (see Table I). These characters have been discussed by Kotov (1999).
The value of a general body shape is similarly compromised in the genera of other families such as Bosmina (Haney and Taylor 2003), where male sexual characters are found to be significantly more valuable (Kotov 1996;Kim et al. 2006). Many species of Daphnia have been established based on inapplicable characters that are too variable within populations and too ecologically plastic to be used (Glagolev 1986a(Glagolev , 1986b. Illusory ''differences'' such as those between females with an empty or full brood pouch in the same species of Moina were specially discussed by Fryer (1991). Only relatively recently have the more applicable ''fine'' characters of appendages armature and secondary male characters been applied to anomopod systematics.
While the antenna I is a good diagnostic character when it is widened distally in maximum expression, it is almost impossible to distinguish ''slightly dilated'' from ''undilated'' (Kotov 1999). All species with a more dilated antenna have been grouped together in the consensus tree, but the probability of this grouping is only 66%. This cannot be regarded as evidence that it appeared only once in the evolution of hirsuticornis-like species. A dilated antenna is characteristic of some other groups of Macrothrix, i.e. M. spinosa and M. laticornis (Smirnov 1976(Smirnov , 1992. It is likely that dilation of antenna II took place independently within different species groups (Kotov 1999). Further, this process is associated with increasing size of some aesthetascs on the tip on antenna I. Indeed, in M. tripectinata and M. flagellata which have a slightly widened antenna, the difference in size between aesthetascs is less than in species with a strongly widened antenna I.
The armature of the largest seta on antenna II (seta on proximal segment of endopod) has traditionally been regarded as helpful for macrothricid systematics (Smirnov 1976(Smirnov , 1992, ever since the first description of Macrothrix triserialis Brady, 1886. Similarly, the armature of lateral setae on segments 1-2 of endopod of antenna II has been found to be very useful for Ilyocryptus (Kotov and Š tifter 2006). Figure 23. A strict consensus of three most-parsimonious trees. The 50% majority rule bootstrap simulation led a tree of similar topology with the consensus tree. Due to this fact, branch probabilities were assigned to the aforementioned consensus tree.
A remarkable hypertrophy of scrapers on limb II of M. boergeni is analogous to that of the distal-most seta on the inner portion of limb III in tropical M. marthae (Elías-Gutiérrez and Smirnov 2000). In both cases, the unusual specialization of the setae, to enable dragging of particles from a substratum, is a reflection of an unusual feeding specialization. Unfortunately, feeding behaviour and other aspects of biology have been investigated only in a few species (Fryer 1974) and remain unknown for most species of Macrothrix.
Unfortunately, ephippial females and males were found only in M. oviformis, so the importance of their characters for species discrimination within the M. hirsuticornis group has not been investigated yet.

Biogeography
Macrothrix hirsuticornis Norman and Brady, 1867 s. str. is absent from the southern hemisphere, where five other, related species occupy different Subantarctic islands and archipelagos, and the southernmost portions of South America and Africa (Figure 24).
M. sarsi sp. nov. is endemic to the Cape region of South Africa, which is regarded as a specific biogeographic zone. Many endemics are confined to the Cape Floral Kingdom (Wishart and Day 2002). The level of endemism in the freshwater fauna of South Africa has been discussed previously (Wishart and Day 2002;Smirnov 2007a). Recent additions (Kotov and Š tifter 2006;Smirnov 2007b) and revised descriptions (Smirnov et al. 2006) may cause our ideas on endemism in South Africa to be reviewed. For example, Lathonura sp., regarded as a Holarctic taxon, was recently recorded in Botswana (Hart and Dumont 2005).
This study increases the number of species recorded from the Antarctic-Subantarctic region (these data have already been incorporated in Table 1 of Forró et al. 2007). It is significant that Macrothrix oviformis together with Daphnia studeri are the only two species of the Cladocera present in continental Antarctica (the former is found only on the Antarctic Peninsula). Even during the Last Glacial Maximum, some refugia of the freshwater fauna existed on the Antarctic continent (Cromer et al. 2006;Gibson and Bayly 2007), so we can assume that these continental populations of Daphnia and Macrothrix were not recently introduced, but existed from a pre-glacial time. At the same time, some other territories, i.e. South Georgia and the South Orkneys, were completely covered with ice and ''must have lost their cladocera completely during the last glacial age'' (Frey 1988). This means that these islands were apparently re-colonized through the agency of migratory birds, or as a result of a passive dispersal of ephippia attached to kelps drifting through sea currents (Akatova 1964). Currently, the main factor in cladocean dispersal has been human activity; for example, Alona weinecki was introduced to Easter Island (Dumont and Martens 1996), where freshwater cladocerans have been long absent.
Only two species of Cladocera can be definitively regarded as widely distributed in the Subantarctic, Alona weinecki and Daphnia studeri (Frey 1988;Pugh et al. 2002;Benzie 2005;Gibson and Bayly 2007). The status of populations of Chydorus cf. sphaericus and Chydorus cf. patagonicus, appearing also to be circum-Subantarctic, must be checked in the future.
Other cladocerans demonstrate other patterns of distribution. Ilyocryptus brevidentatus Ekman, 1905, originally described from the Falklands and South Georgia, and then discovered at Signy Island (Heywood, 1967), was also found in Tierra del Fuego and continental South America (Kotov et al. 2002). Macrothrix oviformis demonstrates a similar distributional pattern, but its range is somewhat wider, and includes also the South Shetland Islands and the Antarctic Peninsula. At present, we can conclude that the Falkland Islands have no known endemic cladocerans (see also Dartnall and Hollwedel 2007). Frey (1993) analysed the distribution of Pleuroxus cf. aduncus in the Subantarctic and the southernmost portion of South America, and found several isolated species. This study also demonstrates that different Subantarctic islands are inhabited by different species of Macrothrix. There is a partial similarity in the distribution of Pleuroxus and Macrothrix species in the Subantarctic, which may reflect the geological and climatic history of this region. Macquarie Island seems to be a most unusual Subantarctic territory. In addition to M. cf. flagellata, absent from all other localities of the South Polar Ocean, the endemic Pleuroxus macquariensis and Daphnia gelida occur there (Dartnall et al. 2005). It is necessary to note that the pattern of global distribution of Macrothrix species, where the number of known taxa in the southern hemisphere is significantly greater than in the northern hemisphere (especially when including the Australian endemics) agrees with the distribution pattern of Pleuroxus species.
There was no evidence of a Gondwanian heritage (Bayly 1995;Pugh et al. 2002;Benzie 2005) in the distribution of Macrothrix. In contrast, South African and South American species are apparently the closest relatives of Laurasian (Palaearctic and, perhaps, Nearctic) M. hirsuticornis s. str. At the same time, M. tripectinata and M. flagellata, two taxa with a more basal position on the phylogenetic tree, are distantly isolated from each other. Most probably, this pattern of distribution appeared as a result of the disruption of a pancontinental (early Mesozoic?) species complex. Locally distributed species from the southern hemisphere can be regarded as relicts of this large Pangaean group, which is in accordance with the recent concept of the cladoceran biogeography proposed by Korovchinsky (2006) and ideas on earlier (Palaeozoic) differentiation of the cladoceran taxa (Sacherová and Hebert 2003;Kotov 2007). In this case, M. sarsi sp. nov. from the Cape region must be regarded as a palaeo-endemic form sensu Harrison (1965).
Only a preliminary revision of the Macrothrix hirsuticornis group was conducted in this study. Macrothrix hirsuticornis s. lat. has been found in other non-Palaearctic localities. The findings from the southern hemisphere are intriguing, especially those from tropical Argentina (Vávra 1900), the Columbian Andes (Stingelin, 1913), and Australia (Smirnov and Timms, 1983). It is likely that the number of hirsuticornis-like species is still underestimated.