Ovalona pascua, Damme, Kay Van, 2016

Damme, Kay Van, 2016, Endemism and long distance dispersal in the waterfleas of Easter Island, Zootaxa 4154 (3), pp. 251-272 : 253-264

Ovalona pascua sp. nov.

( Figs 1–5 )

Alona weinecki Studer, 1878 in Dumont & Martens (1996: pp. 84–91; Figs 1 –35) and subsequent publications on Easter Island ( Dumont et al. 1998; no figures)

not Ovalona weinecki (Studer, 1878) sensu Van Damme & Dumont (2008a: pp. 4–9; Figs 1–2)

Etymology. The epitheton pascua for Easter, as the species is described here from Easter Island. The genus name Ovalona Van Damme & Dumont, 2008 refers to the oval egg-like shape of the habitus in lateral view, combined with the name of the original genus, Alona Baird, 1843 ( Van Damme & Dumont 2008a) . Older O. pascua sp. nov. adult females have a pronounced oval shape in lateral view.

Type locality. Walled pool close to road, North of Hanga Roa village , West of Easter Island, Leg. H.J. Dumont and K. Martens, 18.VIII.1990 (sample # 2 in Dumont & Martens 1996). Detailed coordinates unknown, for Hanga Roa: 27.1500° S, 109.4167° W.

Type material. Holotype (RBIN INV.138001), ovigerous adult parthenogenetic female.

Paratypes (RBIN INV.138002) four adult parthenogenetic females and two ephippia in a separate slide and one ephippial female RBIN INV.138003) and adult male (RBIN INV.138004), each in a slide

Type series in four slides mounted in glycerol-formaldehyde and one tube with wet material in 4% formaldehyde.Wet material one tube from type locality with ca. 50 specimens covering all life stages, in 4% formaldehyde ( RBIN INV.138005). All type specimens deposited at the Royal Belgian Institute for Natural Sciences ( RBIN), Brussels, Belgium .

Additional specimens examined. Ovalona pascua sp. nov. All surface samples from 1987 (H.J. Dumont; HJD), 1990 (HJD, K. Martens; KM) and 1993 (K. Desender; KD) in the Dumont collection at RBIN (Accession Number for the collection is INV IG 32.094) from Easter Island were studied. Specimens were retrieved and studied from the following localities (details after Dumont & Martens 1996): shallow pool in walled compound near Te Ihu o Motu Pare (North of Hanga Roa), Easter Island, Coll. by HJD and KM, 18.VIII.1990 . Large shallow pool, opposite to Pto Espolon (North of Hanga Roa), Easter Island, Coll. by HJD and KM, 18.VIII.1990 . Stone well with drinking water (North of Hanga Roa), Easter Island, Coll. by HJD and KM, 18.VIII.1990 . Stagnant water in cave at Anu Ipu Maengo (North of Hanga Roa), Easter Island, Coll. by HJD and KM, 18.VIII.1990 . Two eutrophic pools East of Maungo Roiho (North of Hanga Roa), Easter Island, Coll. by HJD and KM, 18.VIII.1990 . Stone basin at seashore with seawater seeping in, at Na Manga, Easter Island, Coll. by HJD and KM, 21.VIII.1990 . Eutrophic stone basin used for watering horses at Hanga Tuu Hata opposite Rano Raraku volcano, Coll. by HJD and KM, 21.VIII.1990 . Pool in crevices on basaltic plateau Vai-Atare, SE of Rano Kau, Easter Island, Coll. by HJD and KM, 23.VIII.1990 . Between vegetation in shallow water, Teravaka, Easter Island, Coll. by KD, 20.XII.1993. In reference to the sample numbers listed in Dumont & Martens (1996) for the localities above, animals were found in large numbers and of all life stages (>500) in the 1990 samples #1, 2, 4, 9, 10, 24, 28, 38 collected by HJD and KM and the Teravaka samples collected by KD in 1993. Few (10–50), only parthenogenetic females found in samples #6, 7, 23 and none in #12, 13, 14, 18, 19, 20, 21, 25, 38 collected by HJD and KM in 1990, none in samples from Rano Raraku collected by HJD in 1987. Several samples of the 1990 expedition could not be found, likely in the collection but in other boxes (#3, 5, 11, 15, 17, 18, 26, 29). Additional specimens of O. pascua sp. nov. were deposited under the accession numbers (numbers between brackets refer to sample numbers in Dumont & Martens 1996) RBIN INV.138006 (#4) (wet material, tube) , RBIN INV.138007 (#9) (wet material, tube), RINB INV.138008-10 (#4) (slides, including males in RBIN INV.138009). Chydorus sphaericus (O.F. Müller, 1776) from #2, adult parthenogenetic female, deposited under RBIN INV.138011 (this record cannot be confirmed as only one specimen was found in the samples).

Specimens for comparison. Three adult parthenogenetic females of Ovalona sp., labeled as Alona cambouei from Patagonia, Chile from the Daday Collection, Natural History Museum of Budapest, Hungary . Accession Numbers Cladocera II-537, II-538 and II-661, labeled “ Alona cambouei 1203 -1901 Patagonia – Sylvestri”. Specimens of Ovalona weinecki , loc. freshwater pool, Heard Island , Antarctica, Coll. H.J.G. Dartnall, The Natural History Museum, London (1993.62–71).

Description. Adult parthenogenetic female. General Habitus ( Figs 1 A & 2A).. Length/width ratio ca. 1.65. Habitus oval, strongly convex dorsal region and convex posterior region of the carapace ( Figs 1 A & 2A).

Carapace ( Figs 1 A & 2A) lacking ornamentation. Valve setae ca. 35–50 (ca. 45 on average), in three main groups, long anterior group, medium small group and intermediate sized posterior group starting just beyond the middle. Setae with fine setules and gradually decreasing in size towards the corner. The last seta is relatively long, therefore the row of setae seems to end abruptly, just before the corner, continuing in fine setules on the margin (not groups) (see habitus drawing Fig. 1 A; Fig. 2 in Dumont & Martens 1996). Posteroventral corner without notch and no posterodorsal angle. Ventral margin of the valves straight, slight depression near the middle ( Figs 1 A & 2A). Posteroventral valve corner round. No dorsal keel.

Cephalic structures. Ocellus smaller than eye, equidistant between rostrum and latter ( Figs 1 A & 2A). Well developed rostrum, round and a round even posterior margin (see Dumont & Martens 1996). Apex of the first antenna reaching rostrum, aesthetascs projecting beyond. Head pores ( Fig. 1 E; Dumont & Martens 1996: Fig. 3 and Fig. 24), three main narrowly connected head pores ( Fig. 1 E) all of the same size, close to the broadly rounded round posterior head shield margin ( Dumont & Martens 1996: Fig. 3 ). Postpore distance PP less than one interpore distance IP. Small pores between 3/4th IP ( Fig. 1 E) to one IP distance ( Dumont & Martens 1996: Fig. 24) from the midline and situated at about midway of the main pores ( Fig. 1 E). Labrum ( Fig. 1 B). Keel short in lateral view, quite convex, no lip present. Variation shown in Dumont & Martens (1996), margin mostly straight yet variable, sometimes wavy ( Fig. 1 B). Antennules/A1 ( Fig. 1 C). Short, about 2–2.5 as long as wide. Sensory seta at a third from the antennular apex. Setulated on the margin (four to five rows). Aesthetascs of similar length, two slightly longer, all implanted on the apex. Second antennae/A2 ( Fig. 1 A, 1D). Coxal spine short. Spinal formula 001/101, setal formula 113/003. Seta on the first exopod segment fine and long, reaching beyond the last exopod segment by at least one third to over half its length. Segments ornamented with small denticles (not shown). First endopod segment spine, reaching half of the second endopod segment; terminal spines about the same lengths as the segments they are implanted on. Terminal setae of similar thickness, one terminal exopod seta markedly shorter.

Thoracic structures. Five pairs of thoracic limbs (P1–5). First limb ( Figs 3 A–B). Short round epipodite. First endite with just two setae, second endite with three setae, third endite with four similar setae ( Fig. 3 A). No anterior elements on en1–2. ODL with one slender seta, with short fine setules in distal half and just longer than longest IDL seta ( Fig. 3 B); IDL with two setae implanted with fine setules unilaterally on distal halves, both similar in size and a shorter naked seta about a third to a quarter their length. Short accessory seta on sixth endite, near the base of the ODL (as in Fig. 3 B). About six anterior setule groups with relatively short setules in each group, strongly decreasing in length ventrally. Short setules on the ventral margin. Ejector hooks very narrow and small, subequal. Gnathobase without a seta but with a small tuft of setules ( Fig. 3 A). Second limb/P2 ( Fig. 3 C). Exopodite quadrangular, with a short subapical seta and apical setules. Eight scrapers gradually decreasing in size between the exopodite and the gnathobase. Two first scrapers relatively longest, third slightly shorter than the fourth, sixth to eigth slightly thicker. Setulation of all scrapers uniform, no anterior elements at the base of scrapers. Gnathobase with short tuft of setules, three elements and a filter comb with seven setae of which the first short and modified. Third limb/P3 ( Figs 3 D–H). Epipodite oval-rectangular with short projection. Exopodite ( Figs 3 D–E) with seven setae in 2+5 arrangement. First seta about twice as long as second, both in opposite directions; third seta about twice as long as sixth seta; fourth seta just shorter than sixth, fifth of the same length as the seventh seta and both are half as long as the sixth seta. Fourth seta clearly bent in the middle, oriented away from the epipodite. All setae widely plumose, except for setae six and seven: sixth seta with asymmetric setulation, with strong thicker sparse setules in the second half, seventh seta with short setules. External endite ( Fig. 3 F) with three setae (1’–3’) of which first two quite short, setules not seen and with a small naked element between them, third (3’) shorter and with longer setules; four plumose setae with thicker base on inner side (1”–4”) of same length; one naked element and four small naked setae on internal endite ( Fig. 3 G); the gnathobase ( Fig. 3 H) with a cylindrical sensillum and the typically larger bent seta (plumose) and two elements. Filter comb w ith seven setae ( Fig. 3 F, a–g). Fourth limb / P4 ( Figs 3 I–K). Pre-epipodite round, epipodite oval with short projection. Exopodite round, with six plumose setae; first two exopodite setae of similar lengths, third seta just slightly longer than second; fourth exopodite seta about half as long as the fourth (or just over half), the fifth 1.5 times as long as the fourth; sixth exopodite seta shortest, about half as long as the fifth. Fifth seta just shorter than third seta. All exopodite setae similar except for fifth and sixth, which are narrower; fifth is setulated asymmetrically. Endite ( Figs 3 J–K) with marginal row of four setae (1’–4’), first fine and longer than the flaming torch setae, of which one quite developed and two markedly shorter, each with about five setules ( Fig. 3 J). One elongate naked sensillum ( Fig. 3 J–K, labeled “s”); gnathobase with one long setae which is typically bent over endite and naked projection at its base; on inner side, three plumose setae (1”–3”) gradually increasing in size towards the gnathobase and a small filter comb with five setae. Fifth limb/P5 ( Figs 3 L–M). Pre-epipodite round, with long setules; epipodite oval with short projection. Exopodite ( Fig. 3 L) shape oval, about two times as long as wide, with slightly concave setulated margin between setae three and four; four exopodite setae, first (dorsal) two longest, oriented dorsally; third shorter by one quarter of second exopodite seta, fourth exopodite seta half as long as the second; inner portion of limb ( Fig. 3 M) with ovally elongated inner lobe with long apical setules ( Fig. 3 L); two plumose endite setae (1’–2’), the first reaching over the inner lobe, the second about half as long as the first. Small projection near gnathobase (broad with tuft), but no filter comb or elements ( Fig. 3 M).

Abdominal/Postabdominal structures. Postabdomen ( Figs 1 F). Widest at preanal corner. Dorso-distal angle present, about 90°, length of postabdomen 2–2.5 times as long as wide (measured from deepest point for the width, from the preanal corner to the ventral margin and from the point of implantation of the natatorial setae to the point of implantation of the terminal claw). Even though from measurements, the postabdomen is 2–2.5 times as long as wide, it has a rather elongate shape. Length of the postabdomen and depth of the preanal portion increase in larger adult females (shorter postabdomen in smaller adult females). Ventral and dorsal margins in postanal portion, parallel (to slightly tapering distally). Anal margin about the same length as the postanal margin, and concave, with smooth transition into the postanal margin. Distal margin (dorsodistal angle) slightly or not protruding. Preanal margin straight to convex. Preanal angle well developed, broad and deep, strongly reaching beyond the postanal margin (by about half the terminal claw length). Postabominal armature ( Fig. 1 G). Seven to ten groups of marginal denticles. Proximal groups with three to four denticles in each group, distalmost four to three groups with a larger distal conical denticle and always one to two smaller denticles adjacent to them, or groups with equally sized denticles all the way to the distal corner. These marginal denticle groups increase in size distally, the last two to three (maximally four) distal denticles can be markedly larger, or just gradually increasing. The marginal anal denticles in three to four groups, all fine. The lateral fascicles are fine, with no markedly thick distal spine. Lateral fascicles six to nine postanal groups, each fascicle with about 10–20 spinules, distalmost one relatively thicker. Two distalmost spines of the distal fascicles can reach just beyond the margin, or none. In the anal region, there are two to three parallel rows of lateral fascicles. Terminal claw ( Fig. 1 F). Between one up to 1.5 times as long as the anal margin, not curved strongly but rather straight. Proximal pecten shorter than half the claw length. Basal spine relatively fine and close to the terminal claw along its length; this spine is about a third or just over a third of the terminal claw length. Basal spinules not shown here, but described, together with pectens on terminal claw, in Dumont & Martens (1996).

Male ( Fig. 2 D; Dumont & Martens 1996: Figs 12–13, 19–23 & 26–27). Length to width ratio about 1.75, quite elongate habitus with pronounced deep ventral embayment before the middle ( Fig. 2 D). No ornamentation of the carapace (only “grainy” appearance). Eye and ocellus of approximately the same size, ocellus equidistant between the rostrum and the eye. Marginal setae ca. 40, long in anterior portion, strongly decreasing in size (more than in female) towards the posterior valve corner. Antennules protruding beyond the rostrum. Labral keel with a strong depression ventrally. The male postabdomen ( Figs 2 E & 3A) is two to 2.5 times as long as wide (width in postanal portion), with a strongly truncated broad square distal portion, a generally square shape in the postanal part. Ventral and dorsal distal margin at each side of the claw base of similar length. The gonopore ends just subterminally ( Fig. 3 A). Terminal claw short, as long as or just shorter than the anal margin. Basal spine reaching up to half the claw length and with strong pecten on the dorsal margin. Preanal corner deep, well developed but not as a projection, just as a wide triangle, postanal corner not developed. Marginal spinules ( Fig. 3 A) short and equal, in groups. Lateral fascicles about five to seven groups with long spinules, distal spinule of each group in the last three to four fascicles reaching over the margin. Several rows of lateral fascicles in the anal portion. The first limb/P1 ( Dumont & Martens 1996: Figs 19–22) with a strong, well developed hook consisting of a thick base and a long elongate thin distal portion, about twice as thin as the basal portion ( Dumont & Martens 1996: Figs 21–22). Anterior setules on the limb surface shorter than in females and IDL with two terminal setae with well developed setules ( Dumont & Martens 1996: Figs 19–20).

Ephippial female and ephippium. Ephippial female ( Fig. 2 B) with typically high dorsum, light brown color. Ephippium light to dark brown with a single egg, as for the genus ( Fig. 2 C).

Size. Parthenogenetic female size ranging between 0.31–0.6 mm, average 0.44mm (size distribution in Dumont & Martens 1996). Male size ranging between 0.39–0.41mm ( Dumont & Martens 1996). Ephippial female size distribution in Dumont & Martens (1996), ranging according to the authors, up to 0.6mm in the revised specimens.

Differential diagnosis. Adult parthenogenetic females of Ovalona pascua sp. nov. are distinct by the combination of three connected headpores and no notches on the headshield, an oval-round habitus, a postabdomen with a wide preanal portion, with deep preanal corner and a right dorsodistal angle (i.e., not protruding strongly or not at all). Additional characters are the groups of marginal denticles (about three per group) on the postabdomen, increasing in length distally with two to four larger conical teeth near the distal corner and the seta on the first endopod of the second antenna reaching a third (up to half) its length over the last endopod segment.

For general comparison with other species, see the key and drawings in Sinev (2015). The marginal postabdomen groups of relatively short denticles increasing distally in size (or with two to three distal larger conical teeth), differ from the singular, markedly long teeth with adjacent denticles in most species, as in O. glabra (Sars, 1901) , or even more pronounced, in O. altiplana ( Kotov, Sinev & Berrios, 2010) . It is different from O. archeri (Sars, 1888) , which has larger distal teeth towards the distal end of the postabdomen. The long postabdomen with the angular (but not strongly acute) dorsodistal angle is remotely similar to the widespread O. pulchella- like species like O. pulchella (King, 1853) , O. cambouei (de Guerne & Richard, 1893) , O. glabra (Sars, 1901) and O. bromelicola (Smirnov, 1988) . O. pascua sp. nov. differs clearly from O. cambouei in having connected head pores and from O. bromelicola by the presence of a well developed basal spine on the terminal claw. From O. pulchella , it differs in the shape of the postabdomen, not acutely protruding in O. pascua sp. nov., and in the absence of clear notches on the posterior margin of the head shield. The new species differs from O. glabra and O. pulchella also in the length of the postabdomen. The seta on the first endopod segment of the second antenna is long in O. pascua sp. nov., reaching a third of its length or more over the last endopod segment, while in O. glabra and O. pulchella the seta is short and reaches just beyond this segment. Similarities with O. weinecki (Studer, 1878) are not obvious from the habitus and postabdomen. The postabdomen is completely different in the latter species, with a strongly round distal angle and with smaller marginal denticles ( Fig. 2 F). Finally, O. pascua sp. nov. differs from O. kaingang ( Sousa, Elmoor-Loureiro & Santos, 2015) in postabdomen shape, which is more compact in the latter and in the marginal denticles, which are not groups in O. kaingang but conical serrated teeth (see Sousa et al. 2015).

Distribution and ecology. Easter Island. Localities and map in Dumont & Martens (1996: Map 1), as O. weinecki . Widespread on Easter Island, reported in 51% of the localities in their study (but retrieved here from less localities, see below). In a broad range of habitats. Dumont & Martens (1996) report the species from shallow eutrophic pools, puddles, lake littoral (Rano Raraku), artificial lakelet in a meadow, stone basins (drinking water for cattle and horses), basalt crevices, cave water, wells (up to 6m deep), open water (Rano Kau Lake). Tolerant of brackish water, found in one coastal locality with seawater intrusion, together with amphipods. Also in small waterpockets, e.g., in water in the mouth of a stone statue (“Roca Tallada, Poike Peninsula”).

Ovalona pascua sp. nov. is clearly eurytopic, yet shallow eutrophic temporary freshwater pools can be considered the primary habitat for the species. Re-examination of the samples revealed highest abundance (> 500 specimens /sample) in shallow eutrophic surface pools throughout the island (see under material examined). The proportion of males and ephippial females in these pool habitats is high, with up to as many sexual as asexual stages present. Few (10– 50 specimens) were found in samples from wells, caves and one coastal stone basin with seawater intrusion. In contrast to the report by Dumont & Martens (1996), not a single specimen of O. pascua was retrieved from the samples examined from Rano Raraku and Rano Aroi Lakes. If present at all, the animal seems currently rare in both lakes. However, not all samples from Dumont & Martens (1996) could be studied. From the samples investigated here, it is clear that O. pascua sp. nov. is the most abundant freshwater arthropod, likely the most abundant freshwater invertebrate, on Easter Island. Ostracods are also extremely abundant, yet no single species is as dominant in as many samples as O. pascua sp. nov.

1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 Distribution

pasc PAC* alti NT* kain NT* wein ANT* setu NA

nigr NT* glab NT

arch AU

brom NT

pulc AU OL nura PA*

anam NT* azor PA*

anas PA

camb AT O L cape AT * agua NT* Morphological affinities of Ovalona pascua sp. nov. A simplified character similarity matrix of 12 characters across 17 species is presented in Table 1. All accepted Ovalona species are included, except for O. karelica (Stenroos, 1897) , which remains data deficient. As this species is very different from O. pascua , its absence in the similarity matrix does not affect the discussion. Phylogenetic analysis of these and additional characters resulted in 50% majority rule consensus trees (maximum parsimony) with strong polytomies, except for predicted sibling species, but revealed no deeper groupings in which Ovalona pascua sp. nov. falls (not shown). Instead, the similarity matrix of parsimony informative characters is shown here, which can help as a simple tool to assess potential affinities of O. pascua sp. nov. within the genus. The matrix does not imply evolutionary relationships, yet shows morphological similarities which could be interpreted as homologies. The shape and armature of the postabdomen, the headshield, head pores and limb structures are most indicative of speciation in the genus and are included in the matrix. Autapomorphies not shown, and male characters could not be retrieved for all species and were not shown here. The main character selected by Sinev (2015) to distinguish three species groups in Ovalona , is taken into account as well (postabdomen length) ( Table 1). The similarity matrix may be a useful tool for future description of new chydorid species, as it visualizes potential affinities (or convergences) in a single matrix in a simplified way, from the perspective of the species under consideration.

Sinev (2015) distinguishes three major groups: a small O. nuragica -group and two main larger groups, the O. cambouei- and O. setulosa- group. It is unclear whether these groups should be seen as natural. However this systematic diagnosis allows a quick assignment based on clear character states. Interesting about this grouping, is that New World species with limited distribution ranges all have a short postabdomen lacking a protruding distal margin (except for O. setulosa which has a wide distribution) and that the three species with the largest distribution fall under the longpostabdomen grouping with protruding distal margin (O. pulchella -group). However, it may still group species that are unrelated and which have a convergence in the postabdomen length. According to postabdomen length, Ovalona pascua sp. nov. would fall into the O. setulosa- group as suggested by Sinev (2015), which includes O. anamariae (Sinev & Silva-Briano, 2012) , O. anastasia (Sinev, Alonso, Miracle & Sahuqillo, 2012) , O. azorica (Frenzel & Alonso, 1988) , O. kaingang ( Sousa, Elmoor-Loureiro & Santos, 2015) , O. setulosa (Megard, 1967) and O. weinecki (Studer, 1878) . Likely, this would include the Canadian/Northern US Alona lapidicola Chengalath & Hann, 1981 and Alona borealis Chengalath & Hann, 1981 as well ( Kotov et al. 2015), which remain unassigned to Ovalona . It is worth comparing O. pascua sp. nov. to the O. cambouei- group (in fact, O. pulchella- group), as it is uncertain if postabdomen length can be regarded as a phylogenetically meaningful character. Also, the postabdomen in O. pascua sp. nov. is actually not so short as the measurements (2–2.5 times as long as wide) suggest and when the preanal corner is not as pronounced. Sometimes the shape is a bit more elongate, the preanal corner is not as developed ( Dumont & Martens 1996). However, in younger females (ovigerous, but smaller), the postabdomen may sometimes be quite short (two times as long as wide), even without marked preanal portion.

The O. pulchella- group contains a few species with basically strong similarities in postabdomen shape and limbs, including, at least, O. cambouei (de Guerne & Richard, 1893) , O. pulchella (King, 1853) and O. glabra (Sars, 1901) . I do not consider O. archeri (Sars, 1888) and O. capensis (Rühe, 1914) as part of this group, in particular O. archeri which has special features on the limbs and a completely different male postabdomen, although the postabdomen shape resembles that of O. cambouei . Male postabdomen shape in O. pulchella , O. cambouei and O. glabra in comparison to O. pascua sp. nov., show a basic similarity in the common male postabdomen type in the genus ( Fig. 4 ). The male postabdomen shape shows some affinity with O. glabra , having a broad rectangular distal portion and a short terminal claw ( Fig. 4 B). In this feature, the male is closer to O. glabra than to O. pulchella and O. cambouei , which have a longer terminal claw and a long basal spine ( Sinev 2001a) ( Fig. 4 B). However, O. pascua remains distinct from O. glabra in both male ( Fig. 4 ) and female ( Fig. 5 ) postabdomen. In the male postabdomen, O. pascua retains a deep preanal portion ( Fig. 4 ), as in the females. In addition, O. pascua sp. nov. has very short marginal spinules and lateral fascicles ( Fig. 4 A), versus long marginal and lateral spinules in O. glabra ( Sinev 2001a, 2015).

In the female postabdomen, O. pulchella- related species all have an elongate shape with strongly protruding, acute dorsodistal margin, different from the straight angled, not protruding dorsodistal margin in O. pascua sp. nov. ( Fig. 5 ). The preanal portion in O. pascua sp. nov. is also clearly deeper in comparison to O. pulchella- like species. In addition, O. pulchella and closest relatives have a short seta on the first endopod segment of the second antenna, which seems a valid synapomorphy for this species cluster, while O. pascua sp. nov. has a clearly longer setae here. On the limbs, the flaming torch setae on the fourth endopod lack a distal portion in the O. pulchella- group ( Sinev 2001b), which clearly differs in O. pascua sp. nov. as well.

So, it is clear that O. pascua sp. nov. is morphologically not directly similar to the widespread O. pulchella - cambouei-glabra cluster. In analogy, the postabdomen shape in O. pascua sp. nov. is markedly different from that in O. bromelicola (Smirnov, 1988) and O. karelica , which both have a very short basal spine as well ( Van Damme et al. 2011b, Sinev 2002a).

In the “ setulosa -group” ( Sinev 2015), the postabdomen seems shorter and because of the deep preanal portion, O. pascua sp. nov. might be categorized here. Most species in this group have a rather unpronounced dorsodistal postabdominal corner, which is a rounded or a straight angle, but not strongly protruding as in the O. pulchella - group. However, the setulosa- group sensu Sinev (2015) might not be a natural assemblage. In fact, this seems to include several species that cannot be categorized under an O. pulchella- cluster, with a conspicuous postabdomen. O. weinecki (Studer, 1848) takes a quite isolated position because of its unusual postabdomen, strongly rounded, and with small groups of marginal denticles. In addition, limb characters are quite different (not a bent seta on the exopodite of P3, and no exopodite seta on P2) ( Van Damme & Dumont 2008a, Table 1). Both in female ( Van Damme & Dumont 2008a) and male (Frey 1988) postabdomen shape, O. weinecki is markedly different from O. pascua sp. nov. in having a round distal margin ( Figs 4–5 ). In the similarity matrix ( Table 1), O. weinecki is on the third place, yet in fact, if male features would be included (although not known for all species), the similarity would be less. O. pascua sp. nov. is therefore not identical to O. weinecki , or other “peculiar” species with a round distal margin like O. azorica (Frenzel & Alonso, 1988) (Sinev et al. 2012) and O. capensis (Rühe, 1914) ( Van Damme et al. 2013) .

Out of the 12 selected characters in the matrix, O. pascua sp. nov. shares similarities in 11 characters with O. altiplana Kotov, Sinev & Berrios, 2010 and 10 characters with O. kaingang Sousa, Elmoor-Loureiro & Santos, 2015 ( Table 1). The Easter Island species differs from both in the groups of marginal denticles on the postabdomen, as opposed to long singular teeth (Kotov et al. 2010, Sinev 2015) in O. altiplana and single conical teeth (not groups) in O. kaingang ( Sousa et al. 2015) . In O. kaingang , these teeth are short. Shorter teeth and a distal right angle are also found in O. setulosa (Megard, 1967) ( Sinev 2009) . Both O. kaingang and O. setulosa have a much shorter, more compressed postabdomen than O. pascua sp. nov. Especially the postabdomen shape in O. kaingang is much more compact than in O. pascua sp. nov. and the first endopod seta is shorter. The postabdomen shape of O. altiplana shows a remarkable similarity to that of O. pascua sp. nov., including a deep preanal portion (see Kotov et al. 2010: Fig. 21A). However, the marginal denticles (long teeth in O. altiplana ) and the limb characters differ quite strongly (e.g., relative exopodite setae lengths of third and fourth limbs are completely different) between both species (compare with Kotov et al. 2010).

In the limbs, the third limb (P3) exopodite is perhaps most informative of all limbs in Ovalona . Similarity is clear in the P3 exopodite of most species of the O. pulchella -group, and with O. setulosa and O. kaingang , where the sixth seta (counted from the epipodite towards the endopodite) has a modified setulation and the length of this seta is half the size of the third (longest) seta, in contrast to O. archeri ( Sinev 2002b, 2009, Sousa et al. 2015). Also, the fourth exopodite seta in P 3 in O. pascua is bent, not straight, a feature present in most Ovalona species, including O. setulosa , O. pulchella , O. glabra and O. cambouei , but not in O. archeri or in O. weinecki for example ( Sinev 2001a – b, 2002b, 2009, Van Damme & Dumont 2008a). The function of these modifications is unknown. On the second limb, O. pascua sp. nov. retains a small seta on the exopodite, which in several species is strongly reduced or even absent, as in O. weinecki ( Van Damme & Dumont 2008a) . O. pascua sp. nov. also shares connected head pores with several species, such as O. pulchella , O. glabra , O. altiplana , O. kaingang ( Sinev 2015; Table 1). Many species in the genus have separate pores, a homoplasy in O. nigra (Smirnov, 1996) and in O. cambouei for example. The unrevised taxon Alona patagonica Ekman, 1900 ( Van Damme et al. 2010) ( Ovalona , not Alona ) from the collection of Daday (as Alona cambouei from Patagonia) was studied here as well. Specimens are in a very bad state (hence no figures). It was only possible to investigate some features. The postabdomen has a short angular shape, quite compact (two times as long as wide), and has long, singular marginal denticles (not groups as in O. pascua sp. nov.), somewhat similar to O. altiplana (but the postabdomen in “ patagonica ” is more compact). It remains to be seen if this is a valid species, but it is clearly different from O. pascua sp. nov.

Other unrevised records include “ Alona cambouei ” from Hawaii ( Eldredge et al. 1997, Brehm 1937, Sars 1904, Uéno 1936). These animals also differ from O. pascua sp. nov. in postabdomen. The drawing by Uéno (1936) of a female postabdomen from Mauna Kea, Hawaii, shows an Ovalona with a compact postabdomen and very long, slender marginal denticles as well, again different from O. pascua sp. nov. The Hawaii populations in Uéno (1936) seem to have a straight dorsodistal angle on the postabdomen, not protruding as in O. glabra . Therefore, in addition to assigned and known Ovalona populations, the lesser studied Ovalona from Hawaii and from Patagonia, are also different in morphology from the Easter Island taxon. These populations remain unrevised. Several other species with a rather compact postabdomen, still housed in Alona ( Van Damme et al. 2010) , are also part of Ovalona , such as A. lapidicola Chengalath & Hann, 1981 and A. borealis Chengalath & Hann, 1981 . These species need formal assignment and comparison, however their diagnostic morphologies do not challenge the species status of O. pascua sp. nov.

In overall morphology and across all species, including limb features, O. pascua sp. nov. seems closest to several New World species O. kaingang , O. altiplana and O. setulosa . It is not identical to O. weinecki , differing also strongly in male morphology and seems not directly related to species of the O. pulchella complex, although the male postabdomen shows similarities to that in O. glabra .

Figures

FIGURE 1. Ovalona pascua sp. nov., adult parthenogenetic female, loc. pool North of Hanga Roa village, West of Easter Island, Chile, coll. by H. J. Dumont and K. Martens, 18. VIII. 1990. A, habitus. B, labral keel, lateral view. C, antennule / A 1. D, antenna / A 2. E, head pores. F, postabdomen, lateral view. G, postabdomen, postanal portion, detail with marginal denticles and lateral fascicles.

FIGURE 2. Ovalona pascua sp. nov. (A – E), loc. pool North of Hanga Roa village, West of Easter Island, Chile, coll. by H. J. Dumont and K. Martens, 18. VIII. 1990. Ovalona weinecki (F), loc. freshwater pool, Heard Island, Antarctica, Coll. H. J. G. Dartnall, The Natural History Museum, London (1993.62 – 71). A, Ovalona pascua sp. nov. Adult parthenogenetic female, habitus. B, ephippial female, habitus. C, ephippium with dormant embryo. D, adult male, habitus. E, adult male, postabdomen. F, Ovalona weinecki (Studer, 1878), adult parthenogenetic female, habitus. Scale bars 100 µm (A – D & F) and 50 µm (D).

FIGURE 3. Ovalona pascua sp. nov. Adult parthenogenetic female, loc. pool North of Hanga Roa village, West of Easter Island, Chile, coll. by H. J. Dumont and K. Martens, 18. VIII. 1990. Thoracic structures. A, first Limb / P 1, ODL and IDL omitted. B, outer (ODL) and inner distal lobe (IDL). C, second limb / P 2. D – H, third limb / P 3 exopodite (D – E) and endite (F – H). I – K, fourth Limb / P 4 exopodite (I) and endite (J) with flaming torch setae (K). L – M. fifth limb exopodite (L) and inner portion without setules (M).

FIGURE 4. Ovalona pascua sp. nov. Adult male postabdomen and shape in comparison to selected species (B). A, postabdomen, complete, loc. pool North of Hanga Roa village, West of Easter Island, Chile, coll. by H. J. Dumont and K. Martens, 18. VIII. 1990. B, comparison O. pascua sp. nov adult male postabdomen shape, three specimens from different localities (top, after Dumont & Martens (1996), centre, pool at Hanga Rao, bottom, pool at Hanga Tuu Hata) compared to three species of the circumtropical O. pulchella- complex and to O. weinecki. O. cambouei after Sinev (2015), O. glabra after Sinev (2001 a), O. pulchella after Sinev (2001 b), O. weinecki after Frey (1988). Not to scale (B).

FIGURE 5. Ovalona pascua sp. nov. adult female postabdomen shape (variation, four specimens from different localities, from top to bottom Hanga Tuu Hata, Vai-Atara, Teravaka, Hanga Rao), compared to three species of the circumtropical O. pulchella- complex and to O. weinecki. O. cambouei after Sinev (2001 b), O. glabra after Sinev (2001 a), O. pulchella after Sinev (2001 b) and O. weinecki after Van Damme & Dumont (2008 a). Not to scale.