19 December 1989 PROC. BIOL. SOC. WASH. 102(4), 1989, pp. 894-900 EULIMNADIA OVILUNATA AND E. OVISIMILIS, NEW SPECIES OF CLAM SHRIMPS (CRUSTACEA, BRANCHIOPODA, SPINICAUDATA) FROM SOUTH AMERICA

Joel W. Martin and Denton Belk

Abstract.—Tv^o new species of the clam shrimp genus Eulimnadia are de­ scribed from South America. Eulimnadia ovilunala, a species with spherical eggs bearing circular crater-like depressions, is described from Argentina, and Eulimnadia ovisimilis, a species with stout cylindrical eggs that closely resemble those of the North American species E. texana (Packard), is described from Paraguay.

Species of the clam shrimp genus Eulim­ tion of the junior author in San Antonio, nadia Packard (sometimes considered syn­ Texas; lots from that private collection onymous with Limnadia; see Webb & Bell formed the basis for this report. 1979) in North, Central, and South America The highway designations in Catamarca were the subject of two recent reviews by Province, Argentina, changed between the Belk (1989) and Martin (1989). Although time Arthur Hulse made clam shrimp col­ many traditionally employed characters lections in 1973-1974 and today. We cite were shown by those authors to be variable the current nomenclature according to the and unreliable, an additional character, 14 April 1986 map published by Automovil morphology of the external egg shell, ap­ Club Argentino in addition to including the pears to be conservative and species-spe­ location information originally supplied by cific. In his comparison of described Central Dr. Hulse (in parentheses). and South American species, Martin (1989) examined two South America forms that Eulimnadia ovilunata, new species did not conform to any known species. Figs. 1, 3A-C Those species, described from samples in the collections of Denton Belk in San An­ Eulimnadia sp. A.—Martin, 1989, fig. 5d tonio, Texas, were referred to by Martin as (eggs). Eulimnadia sp. A (DB 305) and Eulimna­ Material.—T>B 303, paratypes, LACM 74- dia sp. B (DB 632). In this paper we describe 107.1, 3 males, 45 females (22 ovigerous), the two new species. Argentina, Catamarca Province, highway 46 W of Andalgala in flood plain of Rio Ama- nao (formerly Route 62, km 1508, at time Materials and Methods of collection), 4 Mar 1974, coll. A. Hulse; Illustrations were made from ethanol- DB 304, paratypes, LACM 74-108.1, 11 fe­ preserved specimens with a Wild M-5APO males (1 ovigerous), Argentina, Catamarca stereoscope with camera lucida. All speci­ Province, highway 46 S of Andalgala (for­ mens are catalogued in the Natural History merly Route 1, km 104, at time of collec­ Museum of Los Angeles County, abbrevi­ tion), 8 Mar 1974, coll. A. Hulse; DB 305, ated LACM. The abbreviation DB refers to holotype female (ovigerous), LACM 73- the cataloging system of the private collec­ 180.1; paratypes, LACM 73-180.2, 19 fe- VOLUME 102, NUMBER 4 895

Fig. 1. Eulimnadia ovilunata: A, Right valve of holotype female; B, Holotype female, right valve removed, with egg enlarged above; C, Frontal region of holotype female; D, Rostrum and eye of one of three poorly preserved males; E, Caudal region of holotype female; F, Caudal region of male. Scale bars indicate 1.0 mm. males (15 ovigerous), Argentina, Catamarca Measurements. —Male carapace 5.5 to 6.0 Province, highway 46 S of Andalgala (for­ mm length, 3.5 to 4.1 mm height (but see merly Route 1, km 45, at time of collection), Remarks below) (n = 3; DB 303, no males 30 Dec 1973, coll. A. Hulse, 1 female de­ in other collections). Female carapace 5.1 stroyed in SEM preparation (Fig. 3). to 6.8 mm length, 3.3 to 4.7 mm height (n 896 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON

Fig. 2. Eulimnadia ovisimilis: A, Carapace (right valve) of holotype female; B, Holotype female with right valve removed; C, Allotype male with right valve removed; D, Head region of holotype female; E, Head region of allotype male; F, Second male clasper of allotype male; G, Caudal region of allotype male; H, Caudal region of holotype female. Scale bars indicate 1.0 mm.

= 76, DB 303, DB 304, DB 305; smallest apaces all shriveled; condition indetermi­ and largest females both ovigerous). nate (see below). Carapace. —Female carapace (Fig. 1 A, B) Head re^zo«.—Female (Fig. IC) with broadly oval, with hinge border domed and broadly triangular rostrum; male (Fig. ID) with three or four lines of growth. Male car­ with very prominent acute rostrum. VOLUME 102, NUMBER 4 897

Fig. 3. Scanning electron micrographs of eggs of Eulimnadia ovilunata (A-C) and E. ovisimilis (D-F): A, Eggs ofE. ovilunata (paratype) from Catamarca, Argentina (DB 305), x 200; B, Single egg of E. ovilunata (from same female as in A), x520; C, Higher magnification of crater-like depression of egg in B showing mound at bottom of depression (arrow), x 1,180; D, Eggs of E. ovisimilis (paratype) from Paraguay (DB 632), x75; E, Cluster of eggs of-E. ovisimilis on epipod of female, x 120; F, Higher magnification of eggs shown in E showing nature of end pieces, x250.

Antennae.—First antennae pseudoseg- on ventral border of each segment; number mented with aesthetascs on anterior border of segments varies from seven to nine. of each lobe. Second antennae natatory, with Male thoracopods.—Not examined (see spines on dorsal border and plumose setae below). 898 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON

Caudal region.—¥ema\e (Fig. IE) and pletely dried out at some time in the past. male (Fig. IF) caudal regions similar, with Structures traditionally described for males 8 to 10 stout downward curved spines on could not be ascertained, and the illustra­ posterior borders and with telsonal fila­ tion of the male rostral region (Fig. ID) ments arising fi-om between second and third should be verified when males in good con­ such spines. dition are collected. Because of the condi­ £•^^5.—Spherical, with oval or circular tion of the males no allotype was designated depressions that appear fringed with small among the three male paratype specimens fingerlike projections of the shell or tertiary in DB 303. envelope, each depression with a slightly convex and relatively smooth floor (Figs. Eulimnadia ovisimilis, new species IB, 3A-C). Figs. 2, 3D-F Type locality. —Argentina, Catamarca, Eulimnadia sp. B.—Martin, 1989, fig. 5e-g highway 46, 45 km S of Andalgala. (eggs). Range.—¥j[\o^n from three localities in Catamarca, Argentina. Material. —DB 632, holotype female Etymology. —From the Latin "luna" (ovigerous), LACM 84-204.1; allotype male, (moon) and "ova" (egg), because the depres­ LACM 84-204.2; paratypes (2 males, 48 fe­ sions on the eggs are reminiscent of craters males (21 of which are ovigerous)), LACM on the lunar surface. 84-204.3, Paraguay, Chaco Departamento, Remarks.—Eulimnadia ovilunata does Parque Nacional Defensores del Chaco, not differ appreciably from many other Tribo Nuevo, "encontrado en regiones ba- species of the genus except by virtue of the jas de laguna recien inundada," from shal­ egg morphology. Of the South American low regions with submerged grass about one species described or reviewed by Martin week after heavy rains filled a formerly dry (1989), only the egg oi E. brasiliensis Sars pond-marsh (Terry Bonace, personal letter is spherical. However, the E. brasiliensis egg to D. Belk), 23 Nov 1984, coll. T. Bonace lacks the minute projections fringing each and D. Drenner. indentation, and the indentations do not ap­ Measurements.—yialQ carapace 6.9 to 7.5 pear as regularly formed and do not bear mm length, 4.5 to 4.7 mm height. Females the clearly defined oval "floor" of the crater (including ovigerous females) 6.0 to 10.0 (compare Martin's (1989) fig. 4c to his fig. mm length, 4.2 to 6.9 mm height. 5d and to Fig. 3A, this paper). The North Carapace. —Female carapace (Fig. 2A, B) American species Eulimnadia diversa Mat- broadly oval, distinctly elevated along dor­ tox, E. agassizii (Packard), and E. antlei sal border, with four lines of growth. Male (Mackin) also have spherical eggs, but the carapace (Fig. 2C) usually smaller, oval but eggs of all three of these species clearly differ not elevated along dorsal border, instead from E. lunaova eggs under high magnifi­ somewhat flattened along hinge line, with cation (see Belk 1989). The eggs oi Eulim­ two or three lines of growth. nadia antillarum (Baird), a species that oc­ Head region. —Female (Fig. 2D) with very curs in North and South America, remain short bluntly rounded rostrum; male (Fig. undescribed, but the caudal region of that 2E) with short but acute rostrum. species is nearly devoid of spines (see Mar­ Antennae. —Similar to that described for tin 1989:fig. 5A), whereas the caudal region E. ovilunata (and many other members of in E. ovilunata bears relatively large, well the genus); second antennae with eight or developed spines in both sexes (Fig. 1E, F). nine segments. The only males in the series are in very Male thoracopods.—Typical for the ge­ poor condition, apparently having com­ nus, with long 2-segmented palp, stout dis- VOLUME 102, NUMBER 4 899 tal spines on clasper border, and small cup­ there is never sexual dimorphism as marked like projection on distal border of clasper as is seen in E. ovisimilis (Sissom 1971, Belk finger (Fig. 2F). 1989). Additionally, the eggs ofE. ovisimilis Caudal region.—SimWax in males (Fig. appear less domed on the end pieces than 2G) and females (Fig. 2H), with 13 to 17 are eggs of E. texana. stout caudal spines and with telsonal fila­ ments arising from about level of fourth Discussion posterior spine. £^fs.—Stout and cylindrical, with par­ Morphology of the external e^ shell has allel grooves separating rounded ridges along been shown to be a useful and reliable char­ cylinder axis and with perpendicular grooves acter in identifying species of the genus Eu­ on end pieces (Fig. 3D-F; see also Martin limnadia (Belk 1989). Some caution is ad­ 1989, fig. 5e-g). vised when consulting previous published Type locality. —Paraguay, Chaco Depart- accounts of egg morphology where scanning amento, Parque Nacional Defensores del electron microscopy was not used. Specifi­ Chaco, Tribo Nuevo, "encontrado en re- cally, some of the illustrations or verbal ac­ giones bajas de laguna recien inundada," counts given by Daday (1926) were shown shallow region with submerged grass. by Belk (1989) to be erroneous or lacking ^an^e—Known only from the type lo­ in sufficient detail to be of taxonomic value, cality. perhaps a result of limitations of the mi­ Etymology. —The name refers to the croscopy available at that time. Even when marked similarity that eggs of this species SEM is applied, however, there may be cases bear to eggs of the North American species where egg morphology will fail to distin­ E. texana (see Belk 1989). guish between two otherwise recognizable Remarks.—The eggs of Eulimnadia ovi­ species. This was pointed out by Mura similis are virtually identical to those of E. (1986) for distinct species of anostracans texana, a predominantly North American that have identical egg morphologies, and species that has been reported, probably the present paper demonstrates that this mistakenly, as far south as Sao Paulo, Brazil scenario also occurs in some species of clam (Lutz 1929, Daday 1926), This is disturbing shrimp (Spinicaudata only; the Laevicau- in light of our recent findings that egg mor­ data do not have sculptured eggs; Martin & phology is often species-specific in the Lim- Belk 1988). This need for caution was noted nadiidae and is in any case a more conser­ by Belk (1989) in anticipation of this sort vative taxonomic character than any of problem. We emphasize again the need previously employed feature. Indeed, we at for using SEM on branchiopod egg mor­ first thought that we had encountered a phology, but advise against using egg mor­ southern hemisphere population of E. tex­ phology to the exclusion of other characters. ana rather than an undescribed form. How­ Finally, as mentioned briefly by Belk ever, there are several salient differences be­ (1989) and Martin (1989), the present status tween the two species. In is. texana males of the genera Limnadia Brongniart and Eu­ and females have a similar rostral mor­ limnadia Packard is unclear. Most mor­ phology. In contrast, sexual dimorphism is phological characters previously used for marked in E. ovisimilis. In females of E. separating the two genera are variable, and ovisimilis the rostrum is rounded and slight­ there is some overlap, leading several work­ ly protruding, whereas in males the rostrum ers to suggest that the two genera should be is attenuated and sharply produced (see Fig. synonymized (Webb & Bell 1979). How­ 2D, E). Although there is known to be vari­ ever, there are at least two characters that ation in rostral morphology in E. texana, serve to separate the two genera for species 900 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON in the Americas. First, the well developed Literature Cited spine on the posteroventral border of the Belk, D. 1989. Identification of species in the con- caudal somite is always easily discerned in chostracan genus Eulimnadia by egg shell mor­ Eulimnadia, whereas in the only species of phology.—Journal of Crustacean Biology 9:115- Limnadia known from the Americas (L. 125. lenticularis) this spine is absent (although a Daday, E. 1926. Monographie systematique des small lobe is present in the same location). Phyllopodes Conchostraces. Troisieme Partle (suite). —Annates des Sciences Naturelles, Zool- Webb & Bell (1979:fig, 1) show a morpho­ ogie, lOe serie, 1926 9:1-81 (505-580). logical gradation from one state to the other Lutz,A. 1929. Dons phyllopodos, observados no Rio based on drawings in existing literature, and Grande do Norte.—Instituto Oswaldo Cruz, suggest that this character is therefore un­ Supplemento das Memorias 5:3-9, 3 plates. reliable. Second, the telsonal filaments arise Martin, J, W. 1989. Eulimnadia belki, a new clam from between the paired spinose postero- shrimp from Cozumel, Mexico (Conchostraca, Limnadiidae), with a review of Central and South caudal borders in all American Eulimnadia, American species of the genus Eulimnadia. — whereas in Limnadia lenticularis the fila­ Journal of Crustacean Biology 9:104-114. ments arise from a location anterodorsal to ,&D.Belk. 1988. A review of the clam shrimp the point where these borders become fused. family Lynceidae Stebbing, 1902 (Branchiopo- We will address these two characters and da: Conchostraca), in the Americas. —Journal of Crustacean Biology 8:451-482. the status of Eulimnadia vs. Limnadia in a Mura, G. 1986. SEM morphological survey on the future paper. egg shell in the Italian anostracans (Crustacea, Branchiopoda).—Hydrobiologia 134:273-286. Acknowledgments Sissom, S. L. 1971. Morphological variation in £«- limnadia texana, Texas' most common eulim- Much of this paper was written or planned nadian Conchostraca.—Texas Journal of Sci­ during a brief visit by JWM to San Antonio, ence 23:295-297. Webb, J. A., & G. D. Bell. 1979. A new species of Texas, during which time the hospitality and Limnadia (Crustacea: Conchostraca) from the assistance of Mary Shug Belk was greatly granite belt in southern Queensland and north- appreciated by both authors. Captain J. em New South Wales.—Proceedings of the Lin- Trent Collier, U.S. Air Force, assisted in the naean Society of New South Wales 103:237- collection of specimens used later for com­ 245. parative work and provided useful discus­ sion. We thank Arthur C. Hulse and Terry (JWM) Life Sciences Division, Natural Bonace for collecting and giving to Denton History Museum of Los Angeles County, Belk the specimens used in these descrip­ 900 Exposition Boulevard, Los Angeles, tions. This work was supported by the Na­ CaUfomia 90007; (DB) Biology Depart­ tional Science Foundation, grant no. BSR- ment, Our Lady of the Lake University of 8615018 to J. W. Martin and L. G. Abele. San Antonio, San Antonio, Texas 78285.