Redescription of Leydigia parva Daday, 1905 and assignment to Parvalona gen. nov. (Cladocera: Anomopoda: Chydoridae)

A morphological investigation of Leydigia parva Daday, 1905 (Chydoridae: Anomopoda: Cladocera), based on specimens from Paraguay (type specimens) and Brazil, clarifies its position in the subfamily and prompts its assignment to a new genus, Parvalona. The affinity of this rare benthic chydorid with Leydigia Kurz, 1875 and Alona Baird, 1843, in which this taxon was placed earlier, is discussed.


Introduction
Leydigia parva Daday, 1905 was described from specimens from two localities in Paraguay; the description (Daday 1905, p 186) was relatively detailed for the standards of that time, and accompanied by drawings. Besides a note by Goulden (1966), who (erroneously) mentioned the species from sediments of Lake Petenxil, Guatemala, it was never recorded again. Frey (1966), followed by Smirnov (1971), assigned this taxon to Alona Baird, 1843, but this giant, polyphyletic genus is currently under revision. Molecular phylogenetic analysis of the Aloninae supports the polyphyly of Alona s.l. as currently understood (Sacherová and Hebert 2003). Several species groups within this assemblage have already been reassigned to new or existing genera, based on thorough morphological research (Dumont and Silva-Briano 2000;Van Damme et al. 2003;Sinev 2004), but a number of further candidates await separation in the future (Kotov and Sanoamuang 2004). Here, we discuss a South American taxon that qualifies for re-evaluation at the genus level, Leydigia parva Daday, 1905. Daday's (1905 slide containing the lectotype and paralectotype is kept in the collection of Daday (DAD) at the Hungarian Museum of Natural History, Budapest; however, the specimens are relatively poorly preserved and not sufficient for redescription. A recent finding of this rare taxon in NE Brazil now offers an opportunity to redescribe it and to elucidate its position within the subfamily Aloninae (Cladocera: Anomopoda: Chydoridae).

Material and methods
A slide with the lectotype and paralectotype was loaned from DAD; these two females could be studied only in toto. In addition, three specimens were picked from samples from a temporary pool in Mandacaru, Lençó is Maranhenses National Park, Maranhao, Brazil, collected by K. Van Damme and Dr D. Van Damme, 16 August 1996, kept at the University of Ghent (UG), Belgium. These animals were placed on slides in a drop of a glycerol-formaldehyde mixture, and two were studied under an optical microscope in toto. Two of the three parthenogenetic females were dissected for analysis of the appendages. Drawings of both A.A.K. (types) and K.V.D. were combined in the figures.
Enumeration of setae and other limb structures is done from the epipodite towards the gnathobase, without any suggestion of homology.

Diagnosis
Parthenogenetic female. Body subovoid, high, postero-dorsal and postero-ventral angle rounded, ventral margin convex in posterior half and straight to concave in anterior half. Rostrum short and relatively blunt, eye and ocellus of similar size. Three major head pores with relatively narrow connection between them, no lateral head pores found. Labrum with trapezium-shaped labral keel, lacking setulation. Valve with numerous, relatively long setae. Postabdomen remarkably widened in comparison to other alonines, with postanal portion four times as long as anal portion, and whole postabdomen larger than head of animal (in lateral view). Preanal margin somewhat longer than anus, straight to depressed, preanal and postanal angle well defined. Each side of postanal portion with row of successive clusters of relatively long marginal denticles. Medially to postanal denticles, numerous lateral groups of remarkably long lateral setules, distalmost setule of each group considerably large. Postabdominal claw approximately as long as preanal margin, or somewhat longer, slightly and evenly curved. Basal spine large, not adpressed to claw.
Antenna I short, with two to three transverse rows of setules at anterior face. Antennular sensory seta arising at distance of one-third of antenna I length from distal end. Nine aesthetascs of different size, longest as long as antenna I, projecting beyond tip of rostrum. Antenna II with branches relatively elongated, all segments cylindrical, with rows of stout setules. Antennal formula (exo/endo): setae 1-1-3/0-0-3, spines 0-0-1/1-0-1. Spine on first (basal) segment of exopod long, reaching or almost reaching tip of second segment.
Limb I without accessory seta, ODL with a long, naked seta, and a rudiment of the second seta. IDL with two relatively large setae of somewhat different size. Endite III with three posterior soft setae and an anterior stiff seta. Endite II with three soft setae of unequal size, anterior seta rudimentary. Endite I with three long soft setae, without a stiff seta. At ventral margin of limb, groups of short denticles. Low maxillar process with a single short seta supplied with a bunch of setules distally. Limb II with exopodite small, lacking setae. Eight scrapers, a series of hillocks posteriorly to scrapers. Gnathobase with prominent basal-ventral angle, distal armature of gnathobase with three elements, filter plate with six setae, two distalmost setae specially short. Limb III with exopodite supplied with two setae of different size distally, and three lateral setae. On inner limb portion, four posterior soft setae. Seven setae in filter plate III. Limb IV with exopodite supplied with six setae not differentiated into lateral and distal group. Three soft setae on inner limb portion. Filter plate with five setae. Limb V with exopodite supplied with a single distal seta and three lateral setae. Gnathobase a greatly reduced, simple projection, no filter plate found.
Size up to 0.60 mm.

Differential diagnosis
Among representatives of the tribe Alonini Dybowski and Grochowski, 1894 sensu Kotov (2000), Parvalona gen. nov. has a unique combination of characters: the postabdomen is conspicuously wide, consisting mainly of a large, evenly curved postanal portion armed with clusters of medium-sized marginal denticles and lateral groups of long, relatively stout setules, limb I with stout denticles instead of long setules on its ventral margin, limb III with unique morphology, with five exopodite setae of which the third is remarkably long and widely spaced from the fourth seta. Characteristic (although not unique) traits are long setules along the ventral margin of the valve, stout setules on segments of antenna II, ODL I with a relatively short seta and a rudimentary second seta, IDL with only two setae. Morphological differences from some closer relatives are reported in Table I.

Amended diagnosis
See diagnosis of the genus.

Redescription
Parthenogenetic female. General: colour (after fixation) colourless to pale yellow (Daday 1905) to reddish brown. In lateral view body subovoid, high (body height/body length50.6520.73 in adults), with maximum height in middle ( Figure 1A-C). In dorsal view, body strongly bilaterally compressed. Dorsal margin regularly arched from tip of rostrum to rounded postero-dorsal angle, posterior margin slightly convex, postero-ventral Limb II: number of setae in filter plate of gnathobase 7-8 7 6 6-7 Limb III: exopodite specially elongated, with two large terminal setae Limb III: number of setae in exopodite 3-7 2? 5 6-7 Limb IV: number of soft setae on inner limb portion 4 ? 3 0-4 Limb V: inner-distal projection with a bunch of strong setules Limb V: number of setae in filter plate 2 ? 0 0-3 margin widely rounded, ventral margin convex in posterior half and straight to concave in anterior half. In contrast to Daday's (1905) description, no wide striation on the carapace was found. Granulate or dotted valves, resulting from internal structures. In the posteroventral portion of the valve, these dots are organized in several bands along the valve margin, as marked by Daday (1905).
Head: relatively small, rounded-triangular in lateral view, with rostrum short and relatively blunt, in contrast to Daday's (1905) description. Also, in Daday's specimens ( Figure 1A) the rostrum is truncated, but it seems to be an artefact of excessive compression of the specimens during slide preparation. Eye and ocellus of similar size, ocellus located approximately in middle of distance from eye to tip of rostrum. Head shield with mandibular articulation of alonine-type (see Frey 1967). Three major head pores ( Figure 2C-E) with a relatively narrow connection between them, central pore of the same size as anterior or posterior one, or somewhat narrower ( Figure 2D, E). Lateral head pores were not seen. In one specimen ( Figure 2D), a clear field around the pores was noted, demarcated by a thin line.
Labrum: labrum with a main body, small distal labral plate, and a large medial labral keel (Figure 2A, B). In lateral view, labral keel trapezium-shaped, without setulation. Ventral portion of labral keel with thickened ridge.  Valves: large, subovoid, with numerous (51; average of two specimens, Figure 1B, C), relatively long setae, located submarginally in posterior portion of margin bases ( Figure 2F). Setae divisible in three groups, of which the most rostral group contains the largest setae, followed by a small group of shorter setae situated ventrally of the region between limbs I and III and a third and final group of larger setae. Posterior margin of valve with a row of numerous setules at some distance from one another, implanted on inner side of carapace ( Figure 2G).
Postabdomen ( Figure 1D-G): postabdomen with remarkably wide postanal portion, at least four times as long as anal margin. Ventral margin straight to slightly convex, with rows of minute marginal setules. Anal margin relatively short, shifted strongly to base of postabdomen. Preanal margin somewhat longer than anus, straight to depressed (there is a chance that it is deformed in Daday's specimens), preanal and postanal angle well defined. Whole postanal margin as large arched curve. Inflated basis for postabdominal claws bordered from postanal margin by a distinct depression. Distally, each side of postabdomen provided with a row of 13-16 successive clusters of relatively long marginal denticles, with size increasing distally and each group consisting of mostly three to five denticles; these clusters continue into three to four groups of fine setules on anal margin. Medially to postanal denticles, 9-13 groups of long lateral setules, the distalmost of each group or row being the largest and thicker than the others.
Postabdominal seta: as long as anal plus preanal margin ( Figure 1D). Postabdominal claw: approximately as long as or little longer than preanal margin, slightly and evenly curved. Basal spine large, as marked by Daday (1905), length up to 1.5-2 times diameter of claw at base; basal spine stout, not pressed to claw. Small basal denticles present, proximally from basal spine ( Figure 2F, G).
Antenna I: short (shorter in lectotype, Figure 2I), not reaching tip of rostrum, with two to three transverse rows of setules at anterior face ( Figure 2H, I). Antennular sensory seta slender, as long as half of antenna I length, arising at distance of one-third of antenna I length from distal end. Nine aesthetascs of different size, longest as long as antenna I, projecting beyond tip of rostrum.
Limb I: epipodite small, globular ( Figure 2N). Accessory seta absent, ODL with a long, naked seta, and a rudiment of the second seta ( Figure 2O, P). IDL of similar size as ODL, with two relatively large setae of somewhat different size, armed with small setules unilaterally in terminal half. Endite 1 with three posterior soft setae ( Figures 2N, O, 3A: 2-4), with size increasing basally, all armed with short hairs bilaterally, and one anterior seta (1) long, stout, with short setules in second distal half. Endite 2 with three soft setae of unequal size (5-7): seta 5 short, armed bilaterally with short setules, seta 6 longest, asymmetrically armed with long setules proximally, seta 5 armed with short setules; two small elements anterior to seta 5, one of which (arrow) clearly a rudiment of a stiff seta on endite 2. Endite 3 with three soft setae (8-10), of which one (8) naked and significantly larger than the rest, two other setae two-segmented, setulated along one side, seta 10 somewhat longer than seta 9. Two ejector hooks (eh) of equal size anteriorly on outer portion of limb corm. Also here, a series of long setules and groups of short denticles ( Figure 3B). Low maxillar process with a single short seta supplied with a bunch of setules distally.
Limb III: epipodite ovoid, exopodite flat, relatively small, with three lateral setae ( Figure 3D-F: 1-3) and two setae of different size distally ( Figure 3D, E: 4-5); exopodite setae 4 and 5 widely spaced from one another, their proximal bases (first fourth of seta) perpendicular to each other. Exopodite setae 1 and 2 sparsely setulated with long filter setules, longest exopodite seta 3 armed with long setules in distal half while proximal half bears shorter setules, pressed to the seta; exopodite seta 4, half the length of exopodite seta 3 ( Figure 3D), bears short setules over complete length, while exopodite seta 5, one-quarter length of previous seta, is only setulated in distal half (note that setulation of exopodite setae can be described better on SEM photographs because of the three-dimensional nature which is represented less in a two-dimensional drawing). Distal endite armed with three stiff setae ( Figure 3D, H: 1-3). Basal endite of similar size with distal endite, anteriorly with a sensillum and four, stiff setae ( Figure 3H: 4-7). Four posterior soft setae (19-49), setulated bilaterally. Distal armature of gnathobase with four setae ( Figure 3G: 1-4), one of them (1) a thick, bottle-shaped sensillum of middle size. Seven setae of similar size in filter plate III.
Ecology. Daday (1905) did not provide details about the localities from Paraguay. The Brazilian specimens were found in a shallow, temporary waterbody (dimensions: 1.5 m6100 m680 m) between cerrado-fixed dunes in the Lencoís Maranhenses, functioning as a drinking pool for cattle. The waterbody was rich in submerged vegetation, with a mean temperature of 31uC, pH 8.4, conductivity of 140 mS cm 21 and oxygen level of 8.4 mg l 21 . Regarding faunistic elements, it was rich in aquatic insects (Hemiptera), molluscs (Ampullaria), frogs and fish. Most common branchiopods consisted of Cyclestheria hislopi Baird, 1859, Leydigiopsis curvirostris Sars, 1901, Alona ossiani Sinev, 1998 (Alona affinis group), Ephemeroporus hybridus (Daday, 1905), and Chydorus ventricosus Daday, 1898. Distribution. This is a rare South American species. At present, it is known only from two(?) localities in Paraguay (Daday 1905) and now a single locality in NE Brazil (first record), both on the Atlantic Coast of the South American continent. It is important to note that although Daday (1905) mentions having studied in his work ''several samples from Paraguay'', we were not able to pinpoint these two locations. The only current locality of which the name resembles Daday's ''Curuzu-chica'' is Curuzú Cuatiá (coordinates 29u489S, 58u029W), a city situated east of the Paraná River in Corrientes Province, Argentina. Situated close to the border of Paraguay, and the Paraná River being a hydrological continuation of the Paraguay River, we believe it possible that this locality might actually be situated in Argentina. Note that all other Central and South American records (Bergamin, 1939;Goulden 1966;Frey 1982) in reality dealt with other species (see ''Discussion'').

A confused synonymy
Our redescription of Parvalona parva comb. nov. makes it clear that previous authors were confused about the identity of this taxon. Goulden's (1966) remains of ''Leydigia parva'' from deposits of Lake Petenxil in Guatemala show a quite distinct armature on the postanal margin of the postabdomen. Smirnov (1971), who placed it in Alona, supposed that the taxon had been described earlier as Birgeia travassosi by Bergamin (1939) and as Alona sp. nov. by Brehm (1939) but we think that this is incorrect. Bergamin (1939) did not actually describe Birgeia travassosi, but offered figures which are quite realistic. The animal depicted has several traits distinguishing it from parva: (1) the body is too high, (2) the rostrum too long and curved, (3) marginal denticles on postabdomen too long, (4) antenna II too long. Birgeia travassosi therefore seems more similar to a juvenile Leydigiopsis curvirostris Sars, 1901 than to Parvalona gen. nov., Alona s.l. or Leydigia. Traits 2 and 3 are also characteristic for Brehm's (1939) material, although the real status of this animal remains unclear. Daday's (1905) original description, here expanded, contains a few characters helpful for the identification of this animal: (1) body egg-shaped, both postero-dorsal and posteroventral angles smooth; (2) rostrum short and pointed; (3) ventral margin of valves slightly convex; (4) short setules along whole ventral margin; (5) dorso-distal angle of postabdomen widely rounded; (6) armature of postanal margin of the postabdomen expressed as a series of setules (''hairs''); (7) series of setules (''hairs'') on sides of postabdomen; (8) anal margin depressed, postanal angle smoothed, while preanal margin prominent and pointed; (9) basal spine on the postabdominal claw well developed; (10) antenna I elongated; (11) relatively small size. Also, we can extract from the author's Figures 20 and 21 that this animal had (12) a large labral keel and (13) a relatively wide postabdomen. Our investigation of the lectotype and paralectotype revealed that suggestions 2, 5, 8, and 10 in Daday's description were inadequate. In reality, the rostrum is relatively blunt; the postanal margin is armed with denticles instead of setules; the postanal angle is well defined; antenna I is relatively short. Helpful for a determination of its taxonomic status are also: (14) three major head pores with a narrow connection; (15) long setules on the body of antenna I; (16) long and stout setules on the basal segment, and on the first and second segments of the exopod of antenna II. Daday's (1905) placement of L. parva was supported only by his suggestion on the ''characteristic shape'' of its postabdomen. In fact, his animal has a series of traits, completely contradicting the diagnosis of the genus Leydigia according to Kotov (2004) (see Table I). In addition, the size of the adult females in Parvalona parva gen. nov. is significantly smaller than that of any Leydigia.
Position of Parvalona gen. nov. within the Aloninae Clearly close to Parvalona gen. nov. in morphology of the postabdomen, is Leydigia glabra, described from Nicaragua by Smirnov et al. (2000). Although the status of this taxon needs to be investigated further, the morphology of the PIII suggests L. glabra to group within Leydigia. As seen in Table I, there are still a number of questions to be answered in this taxon.
Some species groups of Alona Baird, 1843, e.g. members of the A. verrucosa group, a candidate to be lodged in a separate genus (Kotov and Sanoamuang 2004), have relatively long setules in lateral groups on the postabdomen as well, with the distalmost setule of each group conspicuous and strong, as in members of the genus Karualona. However, the postabdomen in these taxa is not as massive and the exopodite of PIII not as elongated, separating them from Parvalona gen. nov. Long spines or spine-like setules on the endopod of the second antenna are also shared by Parvalona gen. nov., Leydigia and members of the Alona verrucosa group. Reduction of setae on the distal lobe of the first limb is also present in members of the Alona verrucosa group where the IDL consists of two setae, but we suggest that this reduction is an independent event in alonine phylogeny. The phylogenetic relationships between these taxa can be clarified once we have a better view of the detailed morphology of other Aloninae, in which a revision of the large and complex genus Alona Baird, 1843 will play a pivotal role. Most important differences between Parvalona gen. nov. and related taxa are given in Table I. The original confusion as to where to place this taxon (first Leydigia, then Alona) is understandable when outer features are considered, but the current study aims to clarify why it belongs to neither.
From a functional-morphological perspective, it is obvious that Parvalona gen. nov. is specialized for life in mud, with adaptations similar to those of the larger benthic inhabitants Leydigia Kurz, 1875 and Leydigiopsis Sars, 1901. Together, these three genera share characters separating them from other chydorids, qualified by Fryer (1968) for Leydigia as specializations for life in muddy substrates: (1) strong bilateral compression of the body; (2) presence of haemoglobin, indicated by a reddish colour (possible induction in Cladocera by environmental factors are not studied); (3) a large and wide postabdomen with strong lateral setules; (4) relatively large marginal setules on the valves; and (5) large exopodite surfaces of limbs IV and V. The specialization of these taxa is most obvious in the morphology of the first, second and third limbs, the main components in handling and scraping detrital particles during benthic feeding (Fryer 1968), and the same can be said for Alona Baird, 1843 as well. At the generic level, food handling in benthic environments is clearly one of the driving forces of alonine adaptive radiation.