I I BULLETIN DE L'INSTITUT ROYAL DES SCIENCES NATURELLES DE BELGIQUE, BIOLOGIE, 66: 73-92, 1996 BULLETIN VAN HET KONINKLIJK BELGISCH INSTITUUT VOOR NATUURWETENSCHAPPEN, B10LOGIE, 66: 73-92, 1996 Redescription and morphological variability of Darwinula stevensoni (BRADY & ROBERTSON, 1870) (Crustacea, Ostracoda) by Giampaolo ROSSETTI & Koen MARTENS Abstract Introduction The species Darwinula stevensoni is extensively redescribed. Mor­ The Darwinuloidea are, together with the Bdelloidea, one phological variability of both valves and soft parts is assessed of the two prominent examples of so-called ancient asex­ in several geographically and climatically distant populations uals (BUTLIN & GRIFFITHS, 1993, JUDSON & NORMARK, and is found to be minimal or non-existant. Only size 1996), organisms which have persisted over geologically significantly varies between populations and this can be long periods of time without sexual reproduction. Accord­ attributed to the differences in ambient temperatures during the larval development. Number and shape of muscle scars also vary, ing to ruling theory, such lineages are doomed to early but this both within and between populations. Valve shape and extinction for several reasons: co-evolving parasites ('Red chaetotaxy of limbs are remarkably constant. One female from Queen' hypothesis of HAMILTON (1980), not VANVALEN, an Italian population has aberrant Mx2-palps, but this specimen 1973), Muller's ratchet and mutation-load-reduction­ is considered a teratological case. theory (MAYNARD SMITH, 1986, JUDSON & NORMARK, Earlier records of males of D. stevensoni and the taxonomic posi­ foe. cit.). Nevertheless, Darwinulidae indeed appear to tion of the infraorder Darwinulocopina within the suborder have survived without sexual reproduction at least since Podocopina are briefly rediscussed. A hypothesis on the the end of the Cretaceous (c. 70 Ma), or the Jurassic (c. biological strategy of darwinulids is tested using data on 120 Ma) (SOHN, 1988), but as quotations of some morphological variability and taxonomic diversity. Mesozoic males are doubtful, possibly since the Permian Key-words: reproduction, diversity, evolution, taxonomy. (250 Ma) (SCHOEN eta!., 1996). It should be noted that 'asexuality' has always equalled to 'absence of males', both Resume in extant populations and in fossil assemblages. A European research network, consisting of six labora­ L'espece Darwinula stevensoni est redecrite en detail. La tories in five countries, has investigated the evolutionary variabilite morphologique, aussi bien des valves que des par­ ecology of reproductive modes of the three extant lineages ties molles, est evaluee dans plusieurs populations geographi­ of non-marine ostracods with a variety of techniques: quement et climatiquement eloignees les unes des autres, et s'est morphological and morphometrical analyses, clonal averee negligeable ou inexistante. Seule Ia taille varie de far;on autecology, starch-gel electrophoresis of allozymes and significative d'une population a !'autre, ce qui peut etre attribue amplification and automated sequencing of both a Ia difference de temperature de !'environment au cours du mitochondrial and nuclear DNA. Three species, one of developpement larvaire. Le nombre et Ia forme des attaches musculaires sont egalement variables, aussi bien, cependant, each lineage, were investigated and Darwinula stevensoni entre differentes populations qu'a l'interieur d'une meme popula­ (BRADY & ROBERTSON, 1870) was the obvious represen­ tion. La forme des valves et Ia chetotaxie des pattes sont remar­ tative of the Darwinuloidea. The present paper forms part quablement stables. Une femelle d'une population italienne of this concerted research effort and reports on mor­ presente d'aberrants palpes Mx2, mais c'est un cas teratologique. phological variability of both valves and soft parts in Les mentions precedentes de males D. stevensoni et Ia position several geographically isolated populations from Europe taxonomique des Darwinulocopina au sein des Podocopina sont and Africa; the clonal and genetic variability of the same reconsiderees. Une hypothese sur Ia strategie biologique des populations will be reported on elsewhere. Darwinula darwinulides est verifiee a I'aide de donnees sur Ia variabilite stevensoni is extensively redescribed. Earlier redescriptions morphologique et Ia diversite taxonomique. Mots-clefs: reproduction, diversite, evolution, taxonomie. of this species (PINTO & SANGUINETTI, 1958, SOHN, 1987) give an incomplete picture of the chaetotaxy and were thus unsatisfactory for our purposes, which included detailed morphological comparisons within and between populations. The position of the Darwinulocopina within the Podocopida is briefly discussed. A hypothesis on the biological strategy of Darwinula is outlined and is tested using the present morphological evidence. 74 G. ROSSETTI & K. MARTENS Materials and methods Hollandersgatkreek, NW Belgium Coordinates: N 51°15'47" E 03°31 '58 " The morphology of both soft parts and valves of Date: 18.5.1995 specimens of four populations (Finland, Belgium, Italy Coli.: K. Martens & M.E. Montenegro and South Africa) was investigated with both Scanning Habitat: slightly saline lake, sandy bottom, animals col­ Electron Microscopy (SEM) and light microscopy. lected at c l-1.5m deep, along edge of littoral Typha-stand. Illustrations of valves and limbs of representative pH= 8.8 specimens are here reproduced. Specimens from the EC = 3.44 mS/cm (S= 2.0) following localities were investigated (Fig. I): water temperature = 12.8 oc Canale Corniano, Italy Lake Paajarvi, Lammi Biological Station, Finland. Coordinates: N 45°1 0'06" E 10°44'44" Coordinates: N 6! 0 03 '02" E 25°02'32" Date : 15.4.1996 Date: 14.11.1995 Coli.: G. Gentile Coli. : K. Martens & K. Salonen Habitat: Material collected in littoral at 1m depth in the Habitat: medium-sized lake, material collected in littoral final stretch of a polluted canal (Canale Corniano) , flow­ at 1-2m deep, on soft mud. No ice at time of collecting. ing into to Lake Superiore of Mantova, a large wetland Station: near field station, stations 'c' and 'e' of RANTA situated along the low course of the Mincio River (a (1979) tributary of the Po River). pH= 7.0 pH= 7.5 EC = 85.8 f.J.S /cm EC = 490 f.l.S/cm water temperature = 4.0 °C. water temperature = 14.3 °C Fig. I.- Location of sampling sites (1: Pii ; 2: Ho ; 3: Os ; 4 : Mo; 5: IM ; 6: CC; 7: Si). I I Morphology of D. stevensoni 75 Lake Sibaya, KwaZulu-Natal, R . South Africa recorded. Size differences among populations were Coordinates: S 27°25'13" E 32°41 '53" analyzed by one-way ANOVA test (SOKAL & ROHLF, Coli.: K. Martens, M. Hamer & M. Coke. 1981 ). Date: 16.10.94 Habitat: oligotrophic lake, situated very near to the ocean, Nomenclature proposed by D ANIELOPOL (1968, 1970) is with sediments generally consisting of clean white sand. used in the description of the chaetotaxy of Antennula Age is presumed Holocene (max 0.0 I Ma). Station: and Antenna and the morphology of the mandibular southernmost arm of lake, at pumping station, opposite masticatory process (Fig. 2). water level gauge jetty. At edge of emergent macrophyte (Typha, Phragmites) stands, amongst submerged macrophytes on very muddy and organically rich sedi­ Abbreviations used in text and figures ment. Max depth of collecting: c. 1 m. pH= 8.6 Soft parts EC = 684 11Sicm A1 antennula water temperature = 22.0 oc A2 antenna Ill incisor on Md coxa Specimens of three other Austrian and Italian populations LaMo lamina molaris are also illustrated (Fig. I) : Md mandibula Mx1 maxilla - Ossiacher See, Austria (N 46°40' E 13°57') (Lab Mx2 maxillula culture). Pg guidance seta ('poi] de guidance') - Lake Montorfano, Italy (N 45°46'57" E 09°08'15"). on Md coxa Sampling date: 25.4.1996. Coli.: V. Rossi & G. Gentile. PI 'poi] laciniforme' on Md coxa - Lake Inferiore of Mantova, Italy (N 45°09'08" RMo molar region on Md coxa E 10°48'21 "). T(l-2) thoracic limbs Sampling date: 11.4.1996. Coli.: G. Rossetti & G. Gentile. Abd abdomen Line drawings of limbs are all made with camera Iucida. Valves Valves are prepared for SEM by washing iil absolute Cp carapace EtOH and simply drying in air. Error on SEM magnifica­ ms muscle scars tions is ±5%. In Darwinula stevensoni often all specimens RV right valve from a preserved recent sample have tightly closed valves, LV left valve which makes dissection without damage to at least one dv dorsal view of the valves very difficult. We developed the following vv ventral view simple technique: glue a piece of double-sided sticky tape llv left lateral view in a small Petri dish and attach the carapace to it. Use IV internal view a very fine brush (no. 000) bearing a small piece of the same tape to push against the edge of the overlapping Sampling sites valve (the R V in D. stevenson f); after a while the valves Pa Lake Paajarvi (Finland) will open slightly. Then open carapace completely with Ho Hollandersgatkreek (Belgium) either brush or fine needle. Add 100% EtOH to the Os Ossiacher See (Austria) mounted carapace in the Petri dish. The valves will come Mo Lake Montorfano (Italy) loose from the tape. Soft parts are then dry and presumed IM Lake Inferiore Mantova (Italy) lost, but can, when needed, be re-soaked again overnight cc Canale Corniano (Italy) in a phosphate rich detergent. Si Lake Sibaya (South Africa) Limbs are prepared for SEM photography in much the same way, namely by submerging and dissecting the body Figured specimens (all adult females) in absolute EtOH and then drying the limbs in air. For limbs, no glue or sticky tape is used, they are directly stuck OC. l817-l 824: Pa; OC.l825- l837: Ho; OC.l838: Os; on a cover slip glued to the stub, and remain attached OC.l839-l845: Mo; OC.l846-l849: IM; OC.l850-1857 : when dried. This has several advantages: small setae and CC; OC.l858-1863 : Si. All specimens stored in the claws cannot sink into the glue and the background of Ostracod Collection of the Royal Belgian Institute of the micrograph is more uniformous.