A phylogenetic analysis of the Conchostraca and Cladocera

Zoological Journal of the Linnean Society (1998), 122: 491–536. With 19 figures

A phylogenetic analysis of the Conchostraca andCladocera (Crustacea, Branchiopoda,Diplostraca)

JØRGEN OLESEN

Zoological Museum, University of Copenhagen, Universitetsparken 15, DK-2100 Copenhagen,Denmark

Received April 1996; accepted for publication October 1996

A computer-assisted cladistic analysis on morphological characters of the Diplostraca (Concho-straca and Cladocera) has been undertaken for the first time. The morphological informationhas been obtained from literature and transformed into 56 suitable characters. The analysisincluded 47 ingroup taxa, comprising five conchostracan taxa (four families of the Spinicaudataand the Laevicaudata) and 42 genera of the Cladocera. A detailed character discussion ispresented which will be a useful working base for future phylogenetic studies on the group.A number of systematic groups were, with differing degrees of certainty, supported in all218 equally short trees. These are the Diplostraca, Cladocera, Gymnomera (Onychopoda andHaplopoda), Onychopoda, Podonidae, Cercopagididae, Anomopoda, Daphniidae, Moininae,Scapholeberinae, Chydoridae, Chydorinae and Sididae. The Spinicaudata were only sup-ported on some of the 218 equally short trees while no support was found for the Conchostraca.Two taxa—the Macrothricidae and Aloninae—were relatively strongly indicated to beparaphyletic. A suggested classificatory hierarchy, without indication of absolute rank, ispresented.

1998 The Linnean Society of London

ADDITIONAL KEY WORDS:—phylogeny – systematics – taxonomy – cladistics – Laevi-caudata – Spinicaudata – Ctenopoda – Anomopoda – Onychopoda – Haplopoda.

CONTENTS

Introduction . . . . . . . . . . . . . . . . . . . . . . . 492Assumptions . . . . . . . . . . . . . . . . . . . . . . . 494

Monophyly or ingroup (Diplostraca) and choice of outgroups . . . . . 494Material and methods . . . . . . . . . . . . . . . . . . . 494Abbreviations . . . . . . . . . . . . . . . . . . . . . . 495Character discussion . . . . . . . . . . . . . . . . . . . . 495Results . . . . . . . . . . . . . . . . . . . . . . . . 524Discussion . . . . . . . . . . . . . . . . . . . . . . . 524

Conchostraca and Cladocera: monophyly and relationships . . . . . . 524The cladoceran orders: Ctenopoda, Onychopoda, Haplopoda and

Anomopoda: monophyly and relationships . . . . . . . . . . 527

Email: [email protected]

4910024–4082/98/040491+46 $25.00/0/zj960096 1998 The Linnean Society of London

J. OLESEN492

The anomopodan families Daphniidae, Macrothricidae, Chydoridae,Bosminidae: monophyly and relationships . . . . . . . . . . . 530

Conclusions . . . . . . . . . . . . . . . . . . . . . . . 531Note added in proof . . . . . . . . . . . . . . . . . . . . 532Acknowledgements . . . . . . . . . . . . . . . . . . . . 532References . . . . . . . . . . . . . . . . . . . . . . . 533Appendix . . . . . . . . . . . . . . . . . . . . . . . . 536

INTRODUCTION

Calman (1909) grouped the branchiopod Crustacea into four orders: AnostracaNotostraca, Conchostraca and Cladocera. This was a slight modification of theearlier schemes put forward by Sars (1867, 1890) and has since been used with onlyfew modifications (see Fryer, 1987a, for a detailed taxomomical review). The mostsignificant modifications were made by Linder (1945) who recognized two distinctivegroups within the Conchostraca (Laevicaudata and Spinicaudata), and later by Fryer(1987a), who elevated the two conchostracan groups and the four groups includedin the Cladocera to ordinal status, in an attempt to emphasize the differencesbetween them.

Gerstaecker (1866–1879) was the first to unite the Conchostraca and theCladocera under the name Diplostraca, and the monophyly of this taxon hasrecently been advocated by Walossek (1993, 1995), who recognized the formationof the bivalved shield (termed ‘secondary shield’) as a defining apomorphy forthe group. The relationship of the higher taxa within the Diplostraca has receivedmuch attention (e.g. Claus, 1876; Eriksson, 1934; Linder, 1945; Schminke, 1981;Fryer, 1987a,b; Martin & Cash-Clark, 1995). An old idea, first proposed byClaus (1876), is that the Cladocera evolved by neoteny from conchostracanswith certain conchostracan larvae as the starting point. Eriksson (1934) exploredthis idea in more detail and was the first to critically evaluate the possiblemonophyletic nature of the Cladocera. His conclusions (reviewed and discussedby Schminke, 1981) were somewhat unclear as he stated both that the individualcladoceran taxa (Ctenopoda, Onychopoda, Haplopoda and Anomopoda) branchedoff from different conchostracan ancestors at various times, and that the Cladocerashould remain as a systematic unit. As Schminke (1981) pointed out, derivingthe Cladocera from the Conchostraca in this way means that the Conchostracaare paraphyletic and the Cladocera polyphyletic. Schminke (1981) furthermoretook up the old idea about the aberrant conchostracan species Cyclestheria hislopiBaird, 1859 (monotypic for Cyclestheriidae) as a ‘link species’ between theConchostraca and the Cladocera. This idea was first proposed by Sars (1887)and has since been repeated, but the phylogenetic consequences have apparentlyonly rarely been discussed in detail (see Martin & Cash-Clark, 1995; Olesen etal., 1997 press for the most recent reviews).

Martin & Cash-Clark (1995) presented, for the first time, a hypothesis for therelationships among the four orders of the Cladocera (Ctenopoda, Anomopoda,Haplopoda and Onychopoda) (Fig. 17B). It was concluded that the Cladocera,despite the large differences between the orders, might still be monophyletic. Theirtheory involved the derivation of the Cladocera from a cyclestheriid-like ancestorbut the relationship of the Cladocera to the Conchostraca was not discussed. Sucha derivation gives Cyclestheria hislopi status as sister group to the Cladocera and wouldimply that the Conchostraca are paraphyletic.

PHYLOGENETIC ANALYSIS OF THE CONCHOSTRACA AND CLADOCERA 493

Wingstrand (1978) presented the first phylogenetic tree involving a major part ofthe taxa within the Diplostraca (Fig. 17A). He based his phylogeny on a combinationof spermatological data and classical morphological characters. The principal resultof the study is that certain spermatological characters (vacuolar maturation typeand heliozoa-like spermatozoa) made Wingstrand (1978) conclude that the cladoceranfamily Macrothricidae may be paraphyletic. These characters appear in taxa outsidethe Macrothricidae and only in some of the macrothricid genera, which implies, ifthe characters evolved only once, that these macrothricid genera are more closelyrelated to groups outside the Macrothricidae than to the rest of the family. Spermmorphology did not contribute to positioning some groups that had traditionallycaused investigators much trouble, such as the two predatory orders Haplopoda(Leptodora Lilljeborg, 1861) and Onychopoda (Polyphemus O. F. Muller, 1785 andother species). Neither could the two families Moinidae (=Moininae) and Bosminidaebe placed.

Major recent contributions towards a phylogenetic understanding of relationshipswithin the Cladocera have been those of Fryer (1963, 1968, 1974, 1991, 1995) who,on the basis of very detailed studies of the functional morphology within thethree largest families of the Anomopoda, discussed different phylogenetic aspects.Moreover, by demonstrating the many differences between the groups included insome of the long established higher taxa of the Branchiopoda, Fryer (1987a,b)suggested that a number of these (including the Diplostraca, Conchostraca andCladocera) should not be considered as monophyletic units and the taxon namestherefore should be abandoned. Unfortunately, he did not at that time propose anysupporting phylogenetic hypothesis. Fryer (1995) gave a hypothesis for the phylogenyof the main lineages within the Anomopoda. While assuming the monophyly ofeach of the four families accepted by him, he stated that the Macrothricidae andDaphniidae are sistergroups as well as the Chydoridae and Bosminidae. He alsogave a detailed account of the presumed structure of the ancestral anomopod.

The very large and diverse anomopod family Chydoridae was split into foursubfamilies, on the basis of head pore morphology (‘neck organ’ or ‘dorsal organ’)and mandible articulation by Frey (1959, 1962, 1967). This was an importantadvance and moreover corresponded to the pattern seen in limb setation (Smirnov,1966). One of the subfamilies, the Aloninae, has recently been suggested to beparaphyletic (Olesen, 1996).

Surprisingly few workers have presented their ideas as testable hypotheses inthe form of cladograms for the Diplostraca. Those that have (Wingstrand, 1978;Schram, 1986; Walossek, 1993; Martin & Cash-Clark, 1995) have beeninconclusive, either because they have only been dealing with higher taxa orbecause of lack of characters. The purpose of the present work is therefore topresent a phylogenetic theory for the Diplostraca (Cladocera and Conchostraca).It is restricted to the Diplostraca because of the almost certain monophyly ofthis group and also because of the challenging variation, especially within theCladocera. Considering the inadequacies in the available phylogenetic theories,it seems useful at this stage to present a phylogenetic hypothesis that combinesthe information provided by earlier workers with some new data, and thatincludes more taxa than in the trees hitherto presented. This paper attempts toresolve the phylogeny within the group as much as possible using the availableinformation and to identify areas where further work is needed. As this is one

J. OLESEN494

of the first cladistic works on diplostracan taxa, it is hoped that it will be auseful starting point for future phylogenetic studies on the group.

ASSUMPTIONS

Monophyly of ingroup (Diplostraca) and choice of outgroups

It is generally accepted that Cladocera and Conchostraca (=Diplostraca) togetherconstitute a monophyletic group (Schram, 1986; Walossek, 1993, 1995; Olesen etal., 1997; see Fryer, 1987a, for a differing view) and it is therefore reasonable to usethis group for a cladistic study. Walossek (1993, 1995) mentions the formation of aunique carapace type (termed ‘secondary shield’, see character 1 in Characterdiscussion below) as a supporting character for the group and Schram (1986)mentions “caudal rami as abreptors” (see character 38 in Character discussion).The presence of natatory antennae has also been used in support of the taxon (e.g.Wingstrand, 1978), but this is most likely a pleisomorphy retained earlier (Fryer,1987a). Olesen et al. (1997) mentions the proximal pair(s) of trunk limbs modifiedas claspers. The ‘secondary shield’ and the ‘claspers’ appear to be the most convincingcharacters since these are present in all species of the Diplostraca and not foundwith this morphology in other taxa.

Furthermore, if the Branchiopoda are a monophyletic group as perceived bysome authors (Sanders, 1963; Wingstrand, 1978; Walossek, 1993, 1995), it thenfollows that the Notostraca and Anostraca would be the closest relatives to theDiplostraca, and are therefore the obvious choices as outgroups. Only a fewcharacters could not be polarized a priori. These include characters related to thecarapace (characters 1–4) and to the claspers (characters 27–29), since these structuresare not present in the outgroups (claspers) or the homologies are uncertain (carapace).

MATERIAL AND METHODS

The characters included in the analysis were defined partly from informationavailable in the literature and partly from morphological studies undertaken inconnection with this work (Olesen, 1996; Olesen et al., 1997). The specific sourcesof literature are in most cases mentioned in the Character discussion. The cladisticanalyses were carried out on a Macintosh platform using the programs PAUP 3.1.1(Swofford, 1993) and MacClade 3.03 (Maddison & Maddison, 1993). The datamatrix appears in the Appendix.

The present analysis does not include all diplostracan taxa. In particular, I haveomitted genera in the two very large cladoceran families, Chydoridae (32 genera)and Macrothricidae (16–17 genera), due to the apparent lack of phylogeneticallyuseful data in the literature. I have been unable to locate characters that were noteither common to the whole family (or a larger assemblage), or unique to the genus.

The material of Cyclestheria hislopi for SEM illustrations was collected in Colombiain 1994. The remaining material for SEM is from the collection of the ZoologicalMuseum, University of Copenhagen. The examined larvae were critical point driedwhile still situated in the maternal brood chamber, dissected free, mounted, gold


coated, and examined with a JEOL JSM-840 scanning electron microscope at 15 kV.Illustrations, redrawn from drawings in literature and in most cases made by

Lillebjorg and Sars, are included to illustrate certain characters (mostly related toexternal morphology), and to give the reader a feeling for the general morphologyof the included taxa. They are not meant to be exhaustive illustrations of charactersor character states; the original literature, mentioned in the character discussion,should be consulted for that purpose.

ABBREVIATIONS

a1 first antenna (antennule)a2 second antennaadf abdominal dorsal filamentbp brood pouchca carapacecau caudal appendagedo dorsal organ (neck organ)exp exopod prolongation (dorsal)fc furcal clawfg food groovego glandular organhs ‘head shield’

la labrummd mandiblemdp mandibular palpmx1 maxilla 1 (maxillule)mx2 maxilla 2 (maxilla)pas post abdominal setaeps proximal setae (antenna 2)rd rectal diverticulumrs rostral spinesf setal fieldt1 trunk limb 1tl trunk limbs

CHARACTER DISCUSSION

This section contains a discussion of the a priori assumptions concerning homologies.

Character 1. Carapace as ‘secondary shield’0=absent1=bivalved, enclosing the whole body2=bivalved, enclosing thorax and abdomen only3=modified to a dorsal brood pouch

Walossek (1993, 1995) interprets the carapace in the Diplostraca (=Onychura) asa ‘secondary shield’ having a different evolutionary and morphological origin fromthe carapace developed to a various degree in a number of other crustaceans(including other branchiopods, like the Notostraca and the Upper Cambrian fossilRehbachiella kinnekullensis Muller, 1983). This interpretation is followed in this treatmentin support for the monophyly of the Diplostraca. At least there is a unique anteriorgrowth of the carapace in the conchostracan taxa, covering the head in the adults(Figs 1, 2).

J. OLESEN496

Figure 1. Three conchostracan species representing 3 of the 5 accepted families. A, female of Lynceusbrachyurus O. F. Muller, 1776 (Lynceidae, Laevicaudata), redrawn after Sars (1896a); B, male ofCyclestheria hislopi Baird, 1859 (Cyclestheriidae, Spinicaudata), redrawn after Olesen et al. (1997); Femaleof Leptestheria siliqva G.O. Sars, 1898 (Leptestheriidae, Spinicaudata), redrawn after Sars (1898). Forabbreviations, see p. 495.

The carapace (‘secondary shield’) has most often been mentioned as starting itsdevelopment from the tergum of the second maxillae segment or the first trunksegment. SEM photographs of developing specimens of Cyclestheria hislopi indicate


Figure 2. Larvae of Cyclestheria hislopi Baird, 1859 (Conchostraca, Spinicaudata) dissected from thefemale brood chamber. All three photographs show larvae belonging to the same stage of development.A, lateral/dorsal view of larva showing the large and globular dorsal organ and the incipient posteriorgrowth of the carapace. The arrow indicates position of the post abdominal setae; B lateral view oflarva. The arrow indicates the anterior margin of the carapace which seems to start development fromthe maxillulary segment; C, ventral view of larva showing the reduced maxillae and the laterallydirected trunk limbs which at this early stage have not started to bend together ventral to the trunk.Scale bars=100 lm.

J. OLESEN498

Figure 3. Leptodora kindti Focke, 1844, the only known species of the Haplopoda (Cladocera). A, adultfemale of Leptodora kindti, redrawn after Sars (1993); B, free-living metanauplius of Leptodora kindti, seenfrom ventral side, redrawn after Sars (1874). Leptodora kindti is the only cladoceran that has retainedfree living larvae as a part of the life cycle (the sexual part).

that the carapace might originate even more anteriorly (Fig. 2b). A clearly bivalvedcarapace is found in most conchostracan and cladoceran genera; in the Conchostracait encloses the whole body (Fig. 1), in the Cladocera only the thorax and abdomen(Figs 7, 9, 10, 12). Exceptions are the two predatory groups, Onychopoda andHaplopoda (Leptodora), where the carapace has been modified into a dorsal broodpouch. Drawings of a Leptodora metanauplius hatched from resting eggs (Sars, 1874),and SEM examination of developing stages of Leptodora (Haplopoda) (Fig. 4B) indicatethat the brood pouch starts developing more anteriorly than the position in theadult would seem to indicate. It is, in fact, more or less in the same position as thecarapace typical of other species of the Conchostraca and Cladocera. This suggeststhat the brood pouch found in Leptodora most likely is homologous with the typicalbivalved carapace.

Character 2. Growth lines of carapace0=without growth lines1=with growth lines


Figure 4. Larvae of Leptodora kindti Focke, 1844 (Cladocera, Haplopoda) dissected from the femalebrood pouch. A, ventral view of larva showing the starting development of the six serially similar trunklimbs, the styliforme mandibles and the prolonged abdomen, all of which are characteristic for theadult; B, dorsal view of a larva at an earlier developmental stage than in ‘A’ showing the dorsal organand the carapace which starts its development more anteriorly than the position in the adult indicates.Scale bars=100 lm.

Growth lines on the carapace shield are found in all species of the Spinicaudata(Conchostraca) (Fig. 1B,C) and in two cladoceran genera, Monospilus G.O. Sars,1862 and Ilyocryptus G.O. Sars, 1862 (Fig. 9B). An undescribed species of theLaevicaudata with growth lines on the carapace was mentioned by Linder (1945)but I assume this is of secondary origin.

J. OLESEN500

A dobp

1

23

4

B

a2

do

a1

cau

pasbpdo

C

Figure 5. Three species of the Onychopoda (Cladocera) representing the three accepted families. A,Bythotrephes longimanus Leydig, 1860 (Cercopagididae), redrawn after Lilljeborg (1901); B, Polyphemusexiguus G. O. Sars, 1897 (Polyphemidae), redrawn after Sars (1902); C, Caspievadne maximovitschi G.O.Sars, 1902 (Podonidae), redrawn after Sars (1902) who described it as belonging to Evadne.

Character 3. Carapace, posterior margin0=not straight and low1=straight and low

A characteristic combination of a strongly curved (sometimes concave) dorsalcarapace margin and a low and straight posterior margin exists in several generaof the Chydoridae (in all genera of Chydorinae). This is illustrated in Sars (1993)and in Lilljeborg (1901) for Chydorus Leach 1816, Pseudochydorus Fryer, 1968, An-chistropus G.O. Sars, 1862, Pleuroxus Baird, 1843, and Alonella G.O. Sars, 1862.

Character 4. Carapace, plumose setae on ventral margin0=setae not plumose1=setae plumose


Figure 6. Larvae of Polyphemus pediculus Linnaeus, 1761 (Cladocera, Onychopoda) dissected from thefemale brood pouch. A, lateral view of late larval stage showing the exopod bearing stenopodouslimbs, characteristic of most onychopods, and the caudal appendage which is shared betweenPolyphemidae and Cercopagididae. The arrow indicates the post abdominal setae; B, dorsal view ofearly larval stage showing the unusual tri-lobed shape of the dorsal organ; C, ventral view of a larvaat same developmental stage as in ‘B’. Scale bars=10 lm.

In two genera, Scapholeberis Schoedler, 1858 and Megafenestra Dumont & Pensaert,1983, rows of unique feather-like setae on the ventral carapace margins can be seen(water-surface association) (Scourfield, 1896; Dumont & Pensaert, 1983). Otheranomopods and the ctenopods have setae on the carapace margins as well, but notmodified this way.

J. OLESEN502

A

a2

ca

do

B

C

a2

la

a1D

Figure 7. Four species of the Ctenopoda (Cladocera) representing the two accepted families. A,Diaphanosoma sarsii Richard (Sididae), redrawn after Sars 1901; B, Holopedium gibberum Zaddach, 1855(Holopedidae), redrawn after Sars (1865); C, Sida crystallina O. F. Muller, 1776 (Sididae), redrawn afterSars (1865); D, Latona setifera O. F. Muller, 1785 (Sididae), redrawn after Sars (1865).

Character 5. Rostral spine0=rostral spine absent1=rostral spine present


Figure 8. Larvae of Ctenopoda (Cladocera) dissected from the female brood chamber. A, ventralview of late larval stage of Diaphanosoma brachyurum Lievin, 1848, showing the six serially similar trunklimbs and the food groove; B, early developmental stage of Diaphanosoma brachyurum Lievin, 1848,showing the incipient development of the carapace and the dorsal organ, the latter of which is lost inthe adult; C, early developmental stage of Sida crystallina O. F. Muller, 1776, showing a typical dorsalorgan which may have some connection to the development of an attachment organ in the adult inthe place of the dorsal organ. Scale bars: A & B, 10 lm; C, 100 lm.

A rostral spine is present in adults of the Leptestheriidae (Fig. 1C) and in juvenilesof at least some species of the Cyzicidae ( juvenile cyzicids with a rostral spine ismentioned by Barnard, 1929).

Character 6. Ventral ganglia0=ventral ganglia coalesced1=ventral ganglia not coalesced

J. OLESEN504

a1

a2

hs

Figure 9. Three species of the Anomopoda (Cladocera) from Bosminidae and ‘Macrothricidae’. A,Bosmina sp. (Bosminidae), redrawn after Lilljeborg (1901); B, Ilyocryptus longiremis G. O. Sars, 1888(‘Macrothricidae’), redrawn after Sars (1901); C, Drepanothrix dentata Euren, 1861 (‘Macrothricidae’),redrawn after Fryer (1974).

In the Haplopoda and Onychopoda the ventral ganglia are more or less coalescedinto a single mass (Calman, 1909), while other branchiopods have a ladder-likenerve-chain.

Character 7. Compound eyes0=eyes not fused1=eyes fused

The compound eyes are fused only in the Cladocera and in the conchostracangenus Cyclestheria G.O. Sars, 1887. No compound eye exists in the chydorid genusMonospilus.

Character 8. Neck organ (dorsal organ), shape0=circular or oval1=elongated in anterior/posterior direction (3 times as long as wide)2=elongated further (4 times as long as wide)


Figure 10. Two species of the Anomopoda (Cladocera) from the family Daphniidae. A, Simocephalustheringi Richard, 1897, redrawn after Sars (1901); B, Moina flexuosa G. O. Sars, 1896 (Moininae),redrawn after Sars (1896b), trunk limb 5 not shown.

3=with median pores4=with median pores, but rim between pores more constricted than in 3)5=median pores with no connection

Many branchiopods and Crustacea in general (see Martin & Laverack, 1992; Martin,1992; Walossek, 1993 for reviews) have a dorsal cuticular ornamentation at theback of the head, referred to as the ‘dorsal organ’, ‘neck organ’ (see position insome taxa in Figs 1–8). The homology between these organs in at least somebranchiopod taxa has been established by Olesen (1996). Olesen (1996) furthermoresuggests a likely transformation series for the neck organ morphology within theChydoridae, used here as an ordered character (see the different neck organ typesin Fig. 18). The character has also been run as unordered.

The following description is from Olesen (1996), unless otherwise stated.In the Eurycercinae, a subfamily of the Chydoridae, and in all examined

branchiopods outside the Chydoridae, the neck organ is circular or oval, or at leastnot of the shape found in the majority of the chydorids (examples of different non-chydorid adult and larval dorsal organs are given in Figs 1–8). In two genera of theChydoridae, Rhynchotalona Norman, 1903 and Tretocephala Frey, 1965, the neck organ

J. OLESEN506

Figure 11. Larval stages of the Anomopoda (Cladocera) dissected from the female brood chamber.A, ventral view of an early stage of Eurycercus glacialis Lilljeborg, 1887 (Chydorinae) showing the initiateddifferentiation of the trunk limbs; B, ventral view of an early stage of Macrothrix laticornis Jurine, 1820(‘Macrothricidae’) showing the large antennules typical for the family, C, ventral view of an early stageof Acantholeberis curvirostris O. F. Muller, 1776 (‘Macrothricidae’); D, ventral view of early stage ofSimocephalus vetulus O. F. Muller, 1776 (Daphniidae). Scale bars: A, 100 lm; B–D, 10 lm.

is elongated in an anterior/posterior direction and is laterally constricted. The neckorgan is most elongated and constricted in Tretocephala.

In a large number of genera in the Aloninae—Leydigia Kurtz, 1874, Alona Baird,


Figure 12. Four species of the Anomopoda (Cladocera) from the family Chydoridae. A, Saycia orbicularisG.O. Sars, 1904 (Saycinae) redrawn after Sars (1904); B, Eurycercus lamellatus O. F. Muller, 1785(Eurycercinae), redrawn after Lilljeborg (1901); C, Alonellanitidula G. O. Sars, 1901 (Chydorinae)redrawn after Sars (1901); D, Camptocercus similis G. O. Sars, 1901 (‘Aloninae’), redrawn after Sars(1901).

1850, Camptocercus Baird, 1843, Acroperus Baird, 1843, Graptoleberis G.O. Sars, 1862and others—the elongated neck organ has been divided into 2–3 median pores,which are still connected by an elevated cuticular rim. The rim between the medianpores is less constricted in Leydigia than in Alona, Camptocercus, Acroperus and inGraptoleberis.

In all genera of the Chydorinae and in two taxa of the Aloninae, OxyurellaDybowski & Grochowski, 1894, and Monospilus dispar G.O. Sars, 1862, the neckorgan consists of one or more median pores, unconnected by any cuticular rim.Monospilus diporus Smirnov & Timms, 1983, an Australian species, displays anarrangement of two interconnected median pores and no lateral pores (light micro-scopy based drawing, Smirnov & Timms, 1983).

Median pores are lacking in Saycia G.O. Sars, 1904 and in other taxa not includedin this analysis (see Frey, 1971, for Saycia).

J. OLESEN508

Figu

re13

.St

rict

cons

ensu

str

eeof

218

equa

llysh

ort

tree

s.C

hara

cter

8w

asin

clud

edas

anor

dere

dch

arac

ter,

all

othe

rch

arac

ters

unor

dere

d.T

hesu

ppor

ted

syst

emat

icgr

oups

are

inbo

ldan

din

clud

eth

eD

iplo

stra

ca,

Cla

doce

ra,

Gym

nom

era,

Ony

chop

oda,

Ano

mop

oda,

Dap

hniid

ae,

Moi

nina

e,Sc

apho

lebe

rina

e,C

hydo

rida

e,C

hydo

rina

e,Si

dida

e.T

axon

nam

esin

italic

sin

dica

tegr

oups

unsu

ppor

ted

byth

isst

udy.

‘1’i

ndic

ates

unsu

ppor

ted

taxa

with

phyl

etic

stat

usun

cert

ain.

‘2’

indi

cate

sta

xasu

gges

ted

tobe

para

phyl

etic

.T

heC

onch

ostr

aca

and

Cte

nopo

daar

eun

supp

orte

dbu

tm

aytu

rnou

tto

besu

ppor

ted

ifch

arac

ters

beco

me

avai

labl

e.T

heM

acro

thri

cida

ean

dA

loni

nae

are,

with

som

ede

gree

ofce

rtai

nty,

indi

cate

dto

bepa

raph

ylet

ic.

(See

Dis

cuss

ion)

.


Figure 14. Strict consensus of 20 equally short trees resulting of successive character weighting inPAUP. Two nodes on Fig. 13 (nodes 1 and 11) are more resolved or differently resolved on this figure.

Figure 15. Majority rule (50%) of the 218 shortest trees. Two nodes on Fig. 13 (nodes 1 and 11) aremore resolved on this figure.

Character 9. Neck organ, lateral pores0=without lateral pores1=with two lateral pores

J. OLESEN510

Figure 16. Strict consensus tree of 10 000 equally short trees. All characters unordered.

Lateral pores, as a part of the neck organ, exist in all species of Eurycercinae,Saycinae and Aloninae (except Monospilus), three subfamilies of the cladoceran familyChydoridae (Frey, 1959, 1962, 1971; Olesen, 1996).

Character 10. Dorsal keel0=dorsal keel absent1=dorsal keel present

In two genera (Camptocercus and Acroperus) of the Aloninae the neck organ is situatedon a large dorsal keel in the head region. The possession of such a keel has earlierbeen mentioned by (Fryer (1968) as a character uniting the two genera (see alsoOlesen, 1996). The conchostracan Cyclestheria hislopi also has the neck organ on akeel (Olesen et al., 1997).

Character 11. Antennular sensillae0=sensillae of antennule not confined to tip1=sensillae of antennule confined to tip

In most genera of Cladocera (Figs 3–12) and in the conchostracan genus Cyclestheria(Fig. 1B), sensillae are found at the tip of the antennule only (female), as for theAnostraca (outgroup).

Character 12. Antennular lobes0=absent1=present


Figure 17. Two cladograms from previous literature that deal with the relationship of the cladoceranmain groups. A, Wingstrand (1978). This phylogeny was mainly constructed on the basis of sper-matological characters, but also a number of other morphological characters were included; B, Martin& Cash-Clark, (1995). A phylogenetic hypothesis for the four main groups of a monophyletic Cladocerawith a cyclestheriid-like ancestor. See comments in text.

J. OLESEN512

Figure 18. Phylogeny of the Chydoridae. The figure is an enlargement of the Chydoridae section inFig. 13 with the different presumed apomorphies listed at left and with the neck organ shown for eachtaxon. The neck organ morphology has been included as an ordered character (8), which is stronglyreflected in the presented phylogeny. The correctness of this transformation assumption can bediscussed, but it is interesting to observe that almost all other located characters either support orsupplement the presumed neck organ transformation. Only character 23—‘number of legs’—conflictedwith the presumed transformation in neck organ morphology (not shown on this figure).


Figure 19. Suggested classificatory hierarchy without specification of absolute rank. Underlined taxaare orders recognised by Fryer (1987a). Only taxa in italic are not supported by this study. ‘1’ indicatesunsupported taxa with phyletic status uncertain. ‘2’ indicates taxa suggested to be paraphyletic.

In all families of the conchostracan order Spinicaudata, except for the Cyclestheriidae,the antennules are long with sensillae-bearing lobes along one side (Fig. 1C).

Character 13. Antennules immobile0=antennules mobile1=antennules immobile

In two genera of the cladoceran order Onychopoda, Podon Lilljeborg, 1853 andEvadne Loven, 1836, the antennules are immobile and fixed to the lower part of thehead (Mordukhai-Boltovskoi, 1968).

Character 14. Male antennule, distal hooks0=absent1=present

J. OLESEN514

In the two genera of the cladoceran subfamily Moininae, Moina Baird, 1850 andMoinodaphnia Herrick, 1887, curved hooks are found in varying numbers on the tipof the long antennule (e.g. Goulden, 1968).

Character 15. Male antennule with one very large seta0=absent1=present

In all genera of the cladoceran order Ctenopoda except Holopedium Zaddach, 1855,the male antennule continues distally into a seta-like projection (e.g. Sars, 1865;Korovchinsky, 1992).

Character 16. Antennal protopod morphology0=protopod not flexed1=protopod flexed

The antennal protopod can be either straight or flexed. A flexed protopod is foundonly in Moininae (Fryer, 1991) and in certain macrothricid genera (in this analysisin Acantholeberis Lilljeborg, 1853, Streblocerus G.O. Sars, 1862, Drepanothrix G.O. Sars,1862 (Fig. 9C), Macrothrix Baird, 1843, Lathonura Lilljeborg, 1853; all according toFryer, 1974).

Character 17. Segmentation of antennal rami0=both rami many segmented1=both rami with 4 segments2=endopod with 3 segments, exopod with 4 segments3=both rami with 3 segments4=endopod with 3 segments, exopod with 2 segments5=endopod with 2 segments, exopod with 3 segments6=both rami with 2 segments

The number of antennal rami segments is very variable but some grouping exists.In the Anostraca (outgroup) the antennae are modified and the number of segmentscannot be counted in the adults. The antennae of the Notostraca (outgroup) areuniramous, reduced or even absent. In the conchostracan order Laevicaudata eachramus consists of approximately 15 segments (Fig. 1A). In the other conchostracanorder, Spinicaudata, there are a little fewer (Fig. 1B,C). In the Daphniidae (includingthe Moininae), the Macrothricidae and the Bosminidae the exopod consists of foursegments, the endopod of three (Figs 9, 10). In the Bosminidae and Ilyocryptus(Macrothricidae) the segments of the exopod are of more or less equal length (Fig.9A,B). In the rest of the above-mentioned groups the most proximal segment isconsiderably shorter than the following ones. In all members of the Chydoridaeboth rami consist of three segments (Fig. 12). In the Ctenopoda some variationexists. In Limnosida G.O. Sars, 1862 (not included), three segments are found in bothrami. In Sida Straus, 1820 there are three segments in the exopod and two in theendopod (Fig. 7B). In Diaphanosoma Fischer, 1850 and Latona Straus, 1820 just theopposite is the case: two segments in the exopod and three segments in the endopod(Fig. 7A,D). In Penilia Dana, 1852 (not included) both rami are two-segmented. In


Holopedium the female antenna is uniramous and two-segmented (Fig. 7C), whereasin the male both rami are two-segmented. In the Onychopoda the endopod is three-segmented and the exopod four-segmented, the proximal segment being considerablyshorter (Fig. 5B,C), exactly like what is found in the Daphniidae (including theMoininae) and most species of the Macrothricidae. In Leptodora each ramus is four-segmented, the proximal segment of the exopod being shorter than the following(Fig. 3A). The ‘exopod’ and ‘endopod’ terminology is mostly from Fryer (1987a).Some states of this multistate character are only found in one taxon in the matrix(states 1,4,6).

Character 18. Antennal muscles0=all extrinsic muscles originate on same side of body as the appendage served1=some extrinsic muscles originate on opposite side of body

An unpublished investigation of G. Fryer (Fryer, 1981a, in part) shows that in theLaevicaudata and in the cladoceran orders some of the extrinsic antennal musclesoriginate on the opposite side of the body as the appendage served. In theSpinicaudata all the extrinsic muscles originate in the same side. Fryer (1987a) statesthat the former situation represents a fundamental evolutionary advance over thelatter. Since I am not aware of precisely which species have been examined, I havechosen to code ‘0’ for the Anostraca, Notostraca and Spinicaudata and ‘1’ for theLaevicaudata and Cladocera.

Character 19. Antennae with two proximal setae0=setae absent1=setae present

Many species in the Anomopoda have a pair of proximal setae at the base of theantennae (Fig. 9B,C). Within the Chydoridae, some genera lack these setae, but thedrawings available in literature may be inaccurate so I have decided to code theseas ‘?’.

Character 20. Mandible articulation0=mandible not articulated on external ‘apodemes’1=mandible articulated on external ‘apodemes’

Within the Chydoridae there are two types of mandibular articulation. In Eu-rycercinae, Saycinae and Aloninae the mandibular articulation is located laterallyat the point where the head shield and carapace unite (fornix). This type ofarticulation appears also to be the situation in the rest of the Branchiopoda and isplesiomorphic for the Diplostraca. In the Chydorinae the mandibles are articulatedon cuticular ‘apodemes’ that carry the points of articulation away from the fornix.The two types of articulation are illustrated and described in detail by Frey (1967),Fryer (1968) and Smirnov (1971).

Character 21. Mandible muscles0=with 5c muscle1=without 5c muscle

J. OLESEN516

The mandibular 5c muscle is missing in two chydorid genera, Anchistropus andPseudochydorus (Fryer, 1968) and in adult Notostraca (Fryer, 1988). According toFryer (1988) such muscles are found in the Anostraca, Spinicaudata, Laevicaudata,Ctenopoda, Anomopoda (except Anchistropus and Pseudochydorus), Onychopoda andHaplopoda.

Character 22. Mandible, anterior process0=mandible without anterior process1=mandible with anterior process

In all examined species of the Onychopoda (except for Polyphemus) an anteriormasticatory mandibular process exists (Martin & Cash-Clark, 1995).

Character 23. Number of trunk limbs0=more than 61=62=53=4

The number of trunk limbs varies from a large number in the Anostraca, theNotostraca and in the two conchostracan orders to 6, 5 or 4 in the cladoceranorders (information mostly from Flossner, 1972).

Character 24. Trunk limb type0=phyllopodous1=stenopodous

Stenopodous trunk limbs are found in two cladoceran orders, the Onychopoda andthe Haplopoda (Figs 3–6).

Character 25. Trunk limbs, serial similarity0=with serial similarity1=without serial similarity

Serial similarity of the trunk limbs is found in the Conchostraca (Fig. 1) and in threeorders within the Cladocera: the Ctenopoda (Figs 7, 8), the Onychopoda (Figs 5, 6)and in the Haplopoda (Figs 3, 4). In the two last-mentioned groups the legs are,however, not of the typical phyllopodous type but stenopodous instead. The serialsimilarity has been lost in the cladoceran order Anomopoda.

Character 26. Trunk limb 1, ejector hooks0=ejector hooks not present1=ejector hooks present

A pair of ‘ejector hooks’ exists near the base of first pair of trunk limbs of all speciesof the Anomopoda. (Fryer, 1987a). These have been described for Eurycercus Baird,


1843 and their function is to remove accumulated detritus under the feeding process(Fryer, 1963).

Character 27. Male clasper with long seta0=long seta absent1=long seta present

Some genera of the Anomopoda have long setae on the male claspers. This is seenin Daphnia O. F. Muller, 1785, Ceriodaphnia Dana, 1853, Metafenestra, Scapholeberis andBosmina Baird, 1845 (Flossner, 1972; Dumont & Pensaert, 1983; Sars, 1993). Thehomologies between these setae are uncertain but are ccoded as the same characterin this analysis.

Character 28. Number of male claspers0=claspers on anterior trunk limbs lacking1=first pair of trunk limbs with claspers2=first and second pair of trunk limbs with claspers

All conchostracan and cladoceran males have the first pair of trunk limbs modifiedto claspers that hold the female during mating (Fig. 1B). Males of the spinicaudateconchostracans (except Cyclestheria) have in addition the second pair of trunk limbsmodified as claspers. Some modifications of the second pair of trunk limbs are seenin some genera of the laevicaudate conchostracans also, but these appear not to beclasper-like (Martin & Belk, 1988). Despite some obvious differences between theclasper morphology in the different taxa, claspers in this treatment are consideredto be homologous structures. The differences between the taxa are thereforeassumed to be secondary modifications of a clasper in the common ancestor oftbe Conchostraca and Cladocera. This homology is not well corroborated bymorphological investigations but it seems, for the moment, to be unnecessarilycomplicated not to assume homology. The clasper morphology is relatively similarin the different taxa (especially in the laevicaudate and spinicaudate conchostracans),and nothing like it is found in the presumed closest relatives to the Conchostracaand Cladocera. Most deviating is Leptodora kindti Focke, 1844 but also in this speciesfirst trunk limbs of the male are apparently modified for clasping. In contrast, themales of the recent Anostraca have the antennae modified for clasping and theDevonian fossil Lepidocaris rhyniensis Scourfield, 1926, with anostracan affinities, hasthe maxillules modified for clasping (Scourfield, 1926, 1940, reinterpreted by Schram,1986 and Walossek, 1993); (See Olesen et al., 1997 for a more detailed discussionand Fryer, 1987a for a differing view).

Character 29. Male claspers, type of palps0=claspers without true palps1=claspers with true palps

According to Botnariuc (1947) and repeated by Fryer (1987a) and Olesen et al.(1997), the so-called palps in the two conchostracan orders, the Spinicaudata andthe Laevicaudata, are not homologous. In the Laevicaudata the palps consist of themodified endites themselves. In the Spinicaudata, including Cyclestheria, the palps

J. OLESEN518

are ‘true’ palps, i.e. outgrowths of the actual endites (see Olesen et al., 1997, for adiscussion of this). The homologies between the conchostracan and cladoceranclaspers with respect to palps are uncertain.

Character 30. Male non-clasper legs0=non-clasper legs without palps1=non-clasper legs with palps

True palps on endites of the male non-clasping legs are found only in the Spinicaudata,except Cyclestheria hislopi (see Olesen et al., 1997, for more details).

Character 31. Trunk limb endites with fork-like setae0=fork-like setae absent1=fork-like setae present

Some macrothricid genera have characteristic fork-like setae on the trunk limbendites. These were first described by Fryer (1974) and are therefore termed ‘Fryer’sforks’ by Smirnov (1992). They are scored ‘present’ in Macrothrix and Streblocerus.

Character 32. Trunk limb 3, distal scrapers0=without distal scrapers1=with distal scrapers

In four genera of the Chydoridae, Camptocercus, Acroperus, Alonopsis G.O. Sars, 1862and Alona, the distal spines of trunk limb 3 serve as scrapers (Fryer, 1968).

Character 33. Trunk limbs 3 and 4, filter plates0=filter plates not enlarged1=filter plates enlarged

Very enlarged gnathobasic filter plates on trunk limb 3 and 4 are found in theDaphniidae (including the Moininae) (Fryer, 1991) (Fig. 10).

Character 34. Trunk limb 5, modification for posterior closing of interlimb space as part offiltration cycle0=not modified1=modified to close

Trunk limb 5 have—in the form of a setose exopod seta—become modified forclosing the posterior interlimb space as a part of the filtering cycle in the Daphniidae(including the Moininae) (Fryer, 1991) (Fig. 10A).

Character 35. Trunk/abdomen joint0=without highly movable joint1=with highly movable joint

Three genera of the Chydoridae, Alonopsis, Acroperus and Camptocercus, have a highly


movable joint between trunk and abdomen. The joint is least flexible in Alonopsisand most flexible in Camptocercus (Fryer, 1968). Some articulation types are illustratedby Smirnov (1971).

Character 36. Abdominal dorsal filaments0=absent1=1 filament present2=2–3 filaments present

Species of the Daphniidae (except Moininae) and the macrothricid genera Ilyocryptushave one or more dorsal filaments extending from the posterior trunk section (Figs9B, 10A). The function is apparently to maintain the embryos within the broodchamber.

Character 37. Paired abdominal setae0=absent1=present

A pair of abdominal setae are found in the Notostraca, Conchostraca and Cladocera,except for the cladoceran Leptodora kindti (Figs 1, 2A, 5, 6A, 7, 9, 10, 11, 12).

Character 38. Furcal claws0=furcal claws not curved1=furcal claws curved

The furcal claws are distinctly posteriorly curved in the Spinicaudata, the Ctenopodaand in the Anomopoda (Figs 1B,C, 7, 9, 10, 12).

Character 39. Long caudal appendage0=long caudal appendage absent1=long caudal appendage present

A prolonged caudal appendage is present in Cercopagididae (Fig. 5A) and Poly-phemidae (Fig. 5B), both of the order Onychopoda. A ‘caudal appendage’ I defineas a very enlarged abdominal/dorsal projection from which the post abdominalsetae arise.

Character 40. Caudal appendage, exuvium retention0=exuvia not retained on the caudal appendage1=exuvia retained on the caudal appendage

In two genera of the Onychopoda (Bythotrephes Leydig, 1860 and Cercopagis G.O.Sars, 1897) the exuvium of the caudal appendage is retained after each moult (Fig.5A).

Character 41. Gut, anterior diverticulae0=present1=present

J. OLESEN520

A gut with anterior diverticulae is found in the Anostraca, the Daphniidae (includingthe Moininae) (Fig. 10B), Holopedium, Eurycercus (Fig. 12b), Ophryoxus G.O. Sars, 1861,and to a lesser extent in Streblocerus (illustrated by Sars, 1865, 1896a, 1993).

Character 42. Gut coiling0=gut not coiled1=Chydoridae coiled2=Eurycercus coiled3=Ophryoxus coiled4=Drepanothrix coiled5=Streblocerus coiled6=Saycia coiled

The gut can be more or less straight or show different loop configurations. Thefollowing different loop types have been recognized, primarily on the basis ofdrawings by Sars (1993) and Lilljeborg (1901). I have chosen not to use a generalcharacter like looped/not looped but have instead looked for similarities betweenthe different loop configuration types. Some loop types are only found in one genus.These are Eurycercus (Fig. 12B), Ophryoxus, Drepanothrix, Streblocerus and Saycia (Fig.12A). One looping type is found in all chydorids (except Eurycercus and in Saycia),where the looping configuration is strikingly similar among all the genera and canbe split into two components: an anteriorly situated dorsally directed loop followedby a posteriorly situated dorsally/anteriorly directed curve (Fig. 12C,D). Somevariation exists but this basic pattern seems always to be present.

Character 43. Gut, rectal dilation/diverticulum0=rectum simple1=rectum dilated2=rectum with diverticulum

A dilation of the rectal gut part is found in the macrothricids Ilyocryptus longiremisG.O. Sars, 1888 (Fig. 9B) and Ophryoxus gracilis G.O. Sars, 1861, and in the chydoridSaycia orbicularis G.O. Sars, 1904 (Fig. 12A) (shown by Sars, 1888, 1901, 1904, 1993).

A rectal diverticulum is found in all included chydorid genera (except Peracantha,see Fryer, 1969 and Saycia, see Sars, 1904), Acantholeberis (Fryer, 1970) and in Ilyocryptussordidus Lievin, 1848 (Sars, 1993). I have chosen to code this as a multistate characterbecause the dilation and the diverticulum apparently appear in the same place ofthe gut. Either a rectal diverticulum or a dilated rectum occur in different speciesof Ilyocryptus, but not in the same species, indicating some connection between thetwo types of gut modifications.

Character 44. Gut, glandular/tubular organs0=organs absent1=glandular organ2=tubular organ

In some genera of the Chydoridae a posterior ventral organ, in most cases anappendix to a rectal diverticulum, is found. The organ is probably an excretory or


osmoregulatory organ and has been investigated by Fryer (1969). He divides theorgan into two different types, a ‘tubular organ’ which he found only in Chydorus,Peracantha, and Pleuroxus, and a ‘glandular organ’ (Fig. 12D) which he found in sevengenera of the Chydoridae. He furthermore considered the organs to be homologousand the tubular organ as the most derived type. Fryer (1969) only had a fewformalin-fixed specimens of Monospilus and Rhynchotalona available but stated that aglandular organ almost certainly is present in both.

Character 45. Position of female genital opening0=ventral1=lateral/dorsal

Included in the group with ‘ventral openings’ are the Anostraca (unpaired openingin a ventral ovisac) and Notostraca/Spinicaudata (openings at the basis of trunklimbs 11) (information from Calman, 1909 and Martin, 1992). According to Linder(1945) the female gonopore in the Laevicaudata is on the basis of trunk limb 11,but later attempts to locate its position have failed (Martin et al., 1986; Martin,1992), and the position is therefore considered as uncertain. In the Cladocera theoviducts open dorsolaterally directly into the dorsal brood chamber (Weismann,1877; Calman, 1909).

Character 46. Number of eggs0=many1=two

The number of parthenogenetic eggs in the cladoceran brood chamber is typicallylarge, but one group, the Chydoridae (except Eurycercus and Saycia), constantly havetwo eggs (Smirnov, 1971) (Fig. 12C,D). The Anostraca, Notostraca and Conchostracaproduce a large number of eggs.

Character 47. Position of eggs (embryos)0=eggs in ventral brood pouch1=eggs encapsulated in exopod2=eggs (embryos) attached to dorsal prolongation of the exopod3=eggs (embryos) brooded dorsally under the carapace

In the Anostraca the 12th pair of limbs are modified into a ventral egg pouch(Walossek, 1993). In the Notostraca the eggs are kept encapsulated in the exopodof trunk limb 11 (e.g. illustrated by Sars, 1899; Fryer, 1988). In both conchostracanorders, the Laevicaudata and Spinicaudata, the eggs (embryos for Cyclestheria hislopi)are attached to a dorsal prolongation of the exopod of certain trunk limbs (Fig. 1A,C) (illustrated by Sars, 1887, 1895; 1896; Martin et al., 1986); trunk limbs 9 and 10in the Laevicaudata and trunk limbs 9–11 in the Spinicaudata (Fryer, 1987a; Martin,1992). In the Cladocera the eggs (embryos) are brooded, non-attached dorsally,between the body and the carapace (Figs 3, 5, 7, 9, 10, 12).

J. OLESEN522

Character 48. Ephippium0=absent1=present

An ephippium (part of the carapace protecting the resting eggs) exists in all generaof the Anomopoda (Flossner, 1972; Fryer, 1972, 1995) and the conchostracan speciesCyclestheria hislopi (Roessler & Sanchez, 1986; Roessler, 1995). The homology betweenthe anomopod and cyclestheriid ephippium is uncertain.

Character 49. Development in brood chamber0=indirect1=direct

The larval development can either be indirect (with free swimming nauplii) or direct(larval stages bypassed before hatching or release from the parent). The indirectdevelopment is found in the Anostraca and Notostraca (outgroups) and in theConchostraca except for Cyclestheria hislopi, where development is direct (Sars, 1887;Roessler & Sanchez, 1986; Roessler, 1995). Direct development in the broodchamber is found in all species of the Cladocera.

Character 50. Development from resting eggs0=indirect1=direct

Development from resting egggs is direct in Cladocera except for Leptodora kindti,where the resting eggs hatch as metanauplii (Sars, 1874) (Fig. 3B). Also in theconchostracan Cyclestheria hislopi the resting eggs sometimes hatch as metanauplii (atleast in populations from Cuba; Botnariuc & Bayes Vina, 1977), which is in contrastto the development in the brood chamber for this species. In Colombian populationsof Cyclestheria hislopi the development from the resting eggs is direct (Roessler, 1995).

Character 51. Nahrboden0=absent1=present

A so-called ‘nahrboden’ (‘placenta’). is found in different taxa. All genera of theOnychopoda, the Moininae and the ctenopod genus Penilia have a nahrboden(Makrushin, 1985; Martin, 1992).

Character 52. Penes0=absent1=present

Penes (paired) are found in two genera of the Ctenopoda, Latona and Diaphanosoma.Penes are also found in the Anostraca and the chydorid Leydigia, but because ofdiffering position are assumed to be non-homologous.


Character 53. Maturation of spermatids0=cystic1=lumenal2=vacuolar3=Bythotrephes type4=Leptodora type5=Latona type6=Polyphemus type7=Podon type8=Evadne type

Wingstrand (1978) categorized the spermatid maturation of most of the investigatedspecies into three different types, which differ with respect to the maturation site:Cystic maturation, lumenal maturation and vacuolar maturation. He found the cystic andthe lumenal types closely related, as in some species of Anostraca a mixture of bothis seen. The vacuolar type he saw as apomorphic and he therefore considered it tobe a character of phylogenetic importance, with parallels only in different mites.This maturation type was found in all investigated species of Anomopoda exceptfor Steblocerus serricaudatus Fischer, 1849 and Ilyocryptus agilis Kurtz, 1878, bothbelonging to Macrothricidae. The spermatogenesis of some species could not becategorized in any of these three groups. This is the case for Leptodora and Latonaand for the species of Onychopoda where, furthermore, great differences betweenthe genera are seen (see Wingstrand, 1978, for description of these maturationtypes). The lumenal type of maturation has been found in Cyclestheria hislopi (Roessler,1995.

Character 54. Spermatozoa, heliozoa-like0=spermatozoa not heliozoa-like1=spermatozoa heliozoa-like

A heliozoa-like sperm type (dark straight rods radiating from the cell centre) is foundin all examined species of Aloninae, Eurycercinae, Simocephalus Schoedler, 1858(Daphniidae), Ophryoxus (Macrothricidae) and in a modified form in two chydorinegenera: Pleuroxus and Peracantha (Wingstrand, 1978). In the phylogeny proposed byWingstrand the presence of this sperm type keeps Ophryoxus together with Daphniidaeand Chydoridae (see Fig. 17a). This implies secondary loss in several genera.Cyclestheria hislopi has not been investigated with TEM, but a light microscopy studyshows that the spermatozoa are not heliozoa-like (Roessler & Sanchez, 1986;Roessler, 1995).

Character 55. Spermatozoa, disc-shaped0=spermatozoa not disc-shaped1=spermatozoa disc-shaped

Disc-shaped spermatozoa occur only in Streblocerus and Ilyocryptus (Macrothricidae)(Wingstrand, 1978). Light microscopy of Cyclestheria hislopi shows that the spermatozoaare not disc-shaped (Roessler & Sanchez, 1986; Roessler, 1995).

J. OLESEN524

Character 56. Spermatozoa, marginal vesicles0=marginal vesicles absent1=marginal vesicles present

Marginal vesicles occur in the spermatozoa of all investigated species of theOnychopoda (Wingstrand, 1978).

RESULTS

Parsimony analysis in PAUP (command: heuristic search, random additionsequence, 1000 replications, TBR swapping), with all characters unordered (exceptcharacter 8 which was ordered), resulted in 218 equally short trees of 121 steps withconsistency and retention indices of 0.72 and 0.92 respectively. The strict consensustree is shown in Figure 13 and the corresponding apomorphy list is shown in Table1. Most characters could be optimized unambiguously on the strict consensus treeof the 218 trees (MacClade command: trace all changes, unambiguous changes only).Characters with ambiguous optimization were optimized ‘manually’ in MacClade, ingeneral preferring ACCTRAN optimization. The 50% Majority Rule consensustree (Fig. 15) had two nodes more resolved than the strict consensus tree (nodes 1and 11 in Fig. 13).

Successive character weighting in PAUP (default settings: best fit of the rescaledconsistency index, and rounded weights) resulted in 20 trees, all found among the218 shortest trees. The strict consensus tree of the 20 trees (Fig. 14) had two nodeseither more or differently resolved than the strict consensus tree of the 218 trees(nodes 1 and 11 in Fig. 13).

A parsimony analysis with all characters unordered (also character 8, which wasunordered in the first analysis) gave thousands of equally short trees of 120 stepswith consistency and retention indices of 0.73 and 0.91 respectively. PAUP was setto retain only 10 000 trees (command: heuristic search, random addition sequence,100 replications, with 100 trees saved for each replication, TBR swapping). Thestrict consensus tree of the 10 000 trees (all characters unordered, Fig. 16) hadconsiderably less resolution than the strict consensus tree of the 218 trees (character8 ordered, Fig. 13). This is due to the loss of some of the neck organ informationgoing from a ordered neck organ character to an unordered neck organ character.Coding character 8 as unordered has the strength that it assumes no a prioritransformation route of the neck organ. The weakness is that large sections of thecladogram remains unresolved.

DISCUSSION

Conchostraca and Cladocera: monophyly and relationships

The Cladocera have in this treatment been supported by a suite of apomorphiesfound in all equally parsimonious trees (characters 1, 17, 23, 45, and 47, see Table1). The majority of these characters are reduction characters (possibly connected toneotenic origin), and the proposed monophyly of the Cladocera may therefore not


T 1. Apomorphy table. The character support for each clade in the strict consensus tree is listed(strict consensus tree. Fig. 13). Possible problems with the supporting character changes have beenindicated by the following numbering system: 1 Reversed later in one or more taxa; 2 Paralleled in oneor more taxa; 3 Reversal; 4 Homologies uncertain; 5 The position of this character depends on how the

node below is resolved

Clade number Character

1 Diplostraca 1 Secondary carapace (0→1)28 One pair of male claspers (0→1)38 Furcal claws, articulated (0→1)1

47 Eggs attached to exopods (1→2)2 Limnadiidae, Leptestheriidae, Cyzicidae 12 Antennular lobes (0→1)

28 Two pairs of male claspers (1→2)30 Palps on male non-clasper legs (0→1)4

3 Leptestheriidae, Cyzicidae 5 Frontal spine (0→1)1

4 Cladocera 1 Carapace reduced to embrace trunk only (1→2)17 Segments of antennal rami reduced to 4 (0→1)23 Number of trunk limbs reduced to 6 (0→1)45 Lateral/dorsal position of gonopore (0→1)47 Eggs/embryos dorsal with no connection to exopod(2→3)

5 Sididae 15 Large seta on male antennule (0→1)6 Latona, Diaphanosoma 52 Penes (0→1)7 Gymnomera 1 Carapace reduced to brood pouch (2→3)

6 Ventral ganglia coalesced (0→1)24 Trunk limbs stenopodous (0→1)

8 Onychopoda 23 Number of trunk limbs reduced to 4 (1→3)51 Nahrboden (0→1)2

56 Spermatozoa with marginal vesicles (0→1)9 Podonidae 13 Antennules immobile (0→1)

10 Cercopagididae 40 Exuvia retained at caudal appendage (0→1)11 Anomopoda 19 Antennae with pair of proximal setae (0→1)

25 Trunk limbs without serial similarity (0→1)4

26 Trunk limb 1 with ejector hooks (0→1)48 Ephippium (0→1)53 Maturation of spermatids vacuolar (0/1→2)1

12 Daphniidae, Macrothrix, Streblocerus, Drepanothrix, 16 Antennal protopod flexed (0→1)1,5

Lathonura 23 Number of trunk limbs reduced to 5 (1→2)1,2

13 Daphniidae 33 Trunk limbs 3 and 4 with enlarged filter plates (0→1)34 Trunk limb 5 modification (0→1)41 Two anterior gut ceacae (0→2)2

14 Scapholeberinae, Ceriodaphnia, Daphnia, 16 Antennal protopod straight (1→0)3

Simocephalus 36 Abdominal dorsal filaments (0→1)2

15 Scapholeberinae, Ceriodaphnia, Daphnia 27 Male clasper with prolonged seta (0→1)1,2

16 Scapholeberinae 4 Ventral carapace margin with plumose setae (0→1)17 Moininae 14 Male antennule with distal hooks (0→1)

51 Nahrboden (0→1)2

53 Spermatid maturation lumenal (2→1)3

18 Macrathrix, Streblocerus 31 Trunk limb endites with fork-like setae (0→1)19 Chydoridae, Ophryoxus 43 Rectal dilation/diverticulum (0→1)2,5

54 Spermatozoa heliozoa-like (0→1)2

20 Chydoridae 9 Neck organ with lateral pores (0→1)1

17 Antennal rami with 3 segments (2→3)21 See Fig. 13 43 Rectal dilation/diverticulum (1→2)2

22 See Fig. 13 8 Neck organ elongated (3 times long as wide) (0→1)44 Glandular organ (0→1)1

46 Two parthogenetic eggs (0→1)23 See Fig. 13 8 Neck organ elongated (4 times long as wide) (1→2)24 See Fig. 13 8 Neck organ with median pores (2→3)25 See Fig. 13 8 Rim between median pores in neck organ constricted

(3→4)

continued

J. OLESEN526

T 1. continued

Clade number Character

26 Alona, Camptocercus, Acroperus, Alonopsis 32 Trunk limb with distal scrapers (0→1)27 Camptocercus, Acroperus, Alonopis 35 Trunk/abdomen with highly movable jint (0→1)28 Camptocercus, Acroperus 10 Neck organ on dorsal keel (0→1)2

29 Chydorinae, Monospilus, Oxyurella 8 Connection between median pores in neck organ lost(4→5)

30 Chydorinae, Monospilus 9 Lateral pores in neck organ lost (1→0)3

31 Chydorinae 3 Carapace with low, straight posterior margin (0→1)20 Mandible articulated on external ‘apodemes’ (0→1)

32 Anchistropus, Pseudochydorus 21 Mandible with 5c muscle missing (0→1)2

33 Pleuroxus, Alonella, Chydorus 44 Tubular organ (1→2)

seem well supported. However, an eventual polyphyletic origin of the Cladoceracan only be shown by identifying potential synapomorphies between separate cla-doceran taxa and other crustaceans—e.g. conchostracans or other branchiopods.Such characters are not present or have not been identified and the Cladocera aretherefore—despite the large differences between the four orders—regarded asmonophyletic until there is evidence to the contrary.

Furthermore, if Cyclestheria is in sister group position to the rest of the spinicaudates(as in Fig. 15), the Cladocera is in addition supported by characters 7 and 49 (‘fusedeyes’ and ‘direct development of parthenogenetic eggs’), which then is paralleled inCyclestheria.

The Conchostraca (Spinicaudata and Laevicaudata) are not supported by anyapomorphic characters. All possible synapomorphies for the Spinicaudata and theLaevicaudata—‘carapace comprising the whole body’ (character 1) and ‘eggs/embryos attached to dorsal prolongation of exopod’ (character 47, see Fig. 1A,C)—turn out most likely to be plesiomorphies already present in the commonancestor to the Diplostraca. I believe that the attachment of eggs to the exopod, asseen in the Conchostraca, is an evolutionary intermediate between ventrally situatedeggs (as seen in the Anostraca and presumably plesiomorphic within the Branch-iopoda) and dorsally situated eggs (as in the Cladocera). The ancestral Cladocera,at some stage, may have had eggs attached to the exopod. Hence, the eggs/embryosattached to a dorsal prolongation of the exopod in the two conchostracan orderscannot be used as an argument for uniting the two groups. Instead, I believe thecharacter is an apomorphy for the common diplostracan ancestory, retained in thetwo conchostracan orders and lost in a cladoceran ancestor, apomorphic to thisgroup. This loss of connection between exopod and eggs could have been coupledto a movement of the gonopores to a more lateral/dorsal position than in otherbranchiopods (character 45), which would facilitate placing the eggs in the dorsalbrood chamber without help from the exopods. Due to the reduced appearance ofthe cladocerans (neoteny), it will probably be difficult to find morphological charactersuniting the conchostracans which could not have been present in the cladoceranancestor. For example, eventual similarities in larval morphology between theconchostracan taxa would be difficult to use convincingly, as in only one case docladocerans have free-living larvae.

The Spinicaudata are only supported on some of the shortest trees (seen in the50% Majority Rule consensus tree, Fig. 15). The supporting characters are: ‘growthlines on carapace’ (character 2, paralleled in Monospilus and Ilyocryptus, see Figs 1B,


C, 9B); ‘claspers with “true” palps’ (character 29, see Olesen et al., 1997) and‘lumenal maturation of spermatids’ (character 53, this maturation type must beassumed to have developed several times). However, a carapace with growth linesmay have been present already in the ancestral diplostracan and the loss of growthlines may therefore represent the apomorphic character state instead of vice versa.

The conchostracan Cyclestheria hislopi (Cyclestheriidae) has often been proposedand discussed as a ‘link’ between the Conchostraca and Cladocera because of certainsimilarities between Cyclestheria and the Cladocera that are not shared with the restof the Conchostraca, e.g. ‘fused compound eyes’, ‘direct development’ and an‘ephippium’ (see Sars, 1887; Linder, 1945; Schminke, 1981; Roessler, 1995; Martin& Cash-Clark, 1995; Olesen et al., 1997). In modern phylogenetic terminology thiswould—if the similarities were interpreted as homologies—give Cyclestheria status assister group to the Cladocera, which would leave the Conchostraca paraphyletic.This position of Cyclestheria is seen in the 20 equally parsimonious trees retainedafter successive character weighting (Fig. 14). Other trees have Cyclestheria in a moreconventional position as sister group to the rest of the spinicaudates which gives amonophyletic Spinicaudata (the supporting characters are mentioned above).

Two possible positions of the Laevicaudata are suggested by the present analysis.A large number of trees have the Laevicaudata as sister group to the Cladocera(seen in the 50% Majority Rule consensus tree, Fig. 15). The support for this restson an unpublished work by G. Fryer (the result is briefly mentioned in Fryer, 1987a),who found a probably derived similarity (=synapomorphy) in muscle arrangementof the antennae in two groups (character 18, see character discussion). If a carapacewith growth lines is plesiomorphic for the Diplostraca, then the loss of growth linescould be seen as further support for grouping the Laevicaudata together with theCladocera. Another possibility is the Laevicaudata as sister group to the rest of theDiplostraca which is seen in the trees retained after successive character weighting(Fig. 14). Support for this is seen in the ‘articulated and backwardly curved furcalclaws’ (character 38), shared by the spinicaudates and two cladoceran orders(Anomopoda and Ctenopoda). In this interpretation the differing furcae of theHaplopoda and Onychopoda have to be assumed to be secondarily achieved. Bothpossible positions of the Laevicaudata leave the Conchostraca paraphyletic.

The ideas above are partly in conflict with those put forward by Fryer (1987a,b),who, while listing the many differences between the cladoceran main groups, rejectsthe Cladocera as a taxonomic unit with phyletic meaning. He also rejects theConchostraca which is in accordance with the view tentatively presented above.

Both the suggested monophyly of the Cladocera and possibly paraphyly of theConchostraca are in agreement with Martin & Cash-Clark (1995) (based on differentarguments), even though the status of the Conchostraca is not stated explicitly inthat paper.

The cladoceran orders Ctenopoda, Onychopoda, Haplopoda and Anomopoda: monophyly andrelationships

The Onychopoda are supported by a number of convincing synapomorphies—‘four trunk limbs’ (character 23, see Figs 5, 6), ‘marginal vesicles in the spermatozoa’(character 56), nahrboden (character 51, paralleled in the Moininae and in thectenopod Penilia)—so there is no serious doubt about the monophyly of this

J. OLESEN528

group. This agrees with the conclusions of Wingstrand (1978), who includes otherspermatozoa-related potential synapomorphies, and with those of Martin & Cash-Clark (1995), who also mention the increase in eye size as a potential synapomorphy.The phylogeny within the Onychopoda I consider as unresolved for the moment.The two families Cercopagididae and Podonidae are each convincingly supportedbut the relationship between these and Polyphemus (Polyphemidae) is more uncertaindue to conflicting characters. A shared anterior process on the mandibles (character22), mentioned by Martin & Cash-Clark (1995), seems to give Podonidae andCercopagididae sistergroup relationship while a shared caudal morphology (caudalappendage, character 39) indicates Polyphemus and Cercopagididae as sistergroups.The latter possibility is supported by a possible apomorphic similarity in the externalmorphology of the dorsal organ/neck organ in Polyphemus and the Cercopagididae(see Figs 5A,B, 6A,B not included as a character in the present analysis). In thesegroups the neck organ is significantly enlarged and not regularly rounded as in thePodonidae and in several other cladocerans (e.g. Macrothrix and Eurycercus, see Olesen,1996).

The Anomopoda, which contains the majority of species within the Branchiopoda,is supported by a suite of apomorphies—the appearance of a ‘unique ejector hookof 1st trunk limb’ (character 26), ‘reduction of serial similarity of trunk limbs’(character 25), by the appearance of ‘ephippium’ (character 48) paralleled inCyclestheria hislopi, ‘antennae with 2 proximal setae’ (character 19, see Fig. 9B,C),‘vacuolar maturation of spermatids’ (character 53) reversed to the cystic maturationtype in Ilyocryptus and Streblocerus and to the lumenal maturation type in the Moininae.These characters are for a large part in accordance with the morphology of theancestral anomopod as convincingly proposed by Fryer (1995).

The analysis yielded no synapomorphies for the Ctenopoda. However, the Sididaeis supported by the ‘special shape of the male antennule’ (character 15). Holopediumdoes not share this morphology and may therefore not belong in the Ctenopoda.Wingstrand (1978) tentatively suggested that Holopedium should be considered assister taxon to the rest of the Cladocera (in the view of Wingstrand, only excludingLeptodora which he placed further down the stem because of the resting eggs hatchingas metanauplii) (Fig. 17A). This possible placement of Holopedium is based on thepresumed primitive presence of simple amoeba-like spermatozoa found in theAnostraca, the Notostraca and in the Conchostraca. The weakness of this placementis that the monophyly of the Cladocera (excluding Holopedium) rests on having a‘non simple amoeba-like spermatozoa’, which include a varied assemblage of differentspermatozoa morphologies.

The relationship among the four cladoceran orders has not been clarified by thisstudy. The results support the Onychopoda and Haplopoda (Leptodora kindti) as amonophyletic group (the name would be the Gymnomera), based on the more orless similar ‘stenopodous trunk limbs’ (character 24), the modification of the carapaceto a ‘brood pouch’ (character 1) (see Figs 3–6), and the ‘coalesced ventral ganglia’(character 6). This view is shared by Martin & Cash-Clark (1995), who also includeother characters like ‘loss of epipods’, ‘loss of food groove’, ‘loss of ocellus’. Thisderivation of Leptodora kindti preserves metanaupliar hatching from the resting eggs(character 50), which forces ‘direct development in the resting eggs’ to be developed2–4 times within the Cladocera.

Many authors (Eriksson, 1934; Wesenberg-Lund, 1952; Fryer, 1987b) havesuggested that the similarities between the two gymnomeran groups (Haplopoda


and Onychopoda) have been convergently derived. Wesenberg-Lund saw Leptodorakindti as closely related to the Ctenopoda and the Onychopoda as closely related toAnomopoda. It is interesting to observe a detailed resemblance between the maleantennule of Leptodora and the antennule of certain ctenopods like Latona (illustratedin Fig. 7D). Wingstrand (1978) suggested a sister group relationship between Leptodoraand the rest of the Cladocera, based on the retention of the free-living larvae (Fig.17A). This is also reflected in the taxonomic scheme of Eriksson (1934), whereLeptodora were given an isolated position in the Cladocera. It would however, involveconsiderable difficulty to derive the Haplopoda and the Onychopoda separately(convergent development of the stenopodous legs and the brood pouch), so Gym-nomera (Haplopoda and Onychopoda) is preferred as monophyletic until newevidence points in another direction.

My analysis does not resolve the relationship between the Gymnomera (Haplopodaand Onychopoda), the Ctenopoda and the Anomopoda (Figs 13–16). Martin &Cash-Clark (1995) proposed the Gymnomera and Anomopoda as sister groups withthe Ctenopoda as sister group to these (Fig. 17B). That may be the correctrelationship, but several problems are connected to the four characters supporting theGymnomera/Anomopoda clade. The first character, ‘reduction of serial similarity’(character 25 in this treatment), may not be a synapomorphy of Gymnomera andAnomopoda but instead an autapomorphy for the Anomopoda, since serial similarityis still found in the Gymnomera, only the legs being stenopodous instead of typicallyphyllopodous (Figs 3–6). The second character, ‘reduction in size of antennules’, isnot present in the Macrothricidae (Fig. 9B,C), a very large family within theAnomopoda. The antennules would then have to be assumed secondarily enlarged.The third character is ‘loss or reduction of metachronal beat’. It is true that noneof the groups have the typical metachronal beat by phyllopodous limbs similar tothose found in anostracans, notostracans, conchostracans and in the cladocerangroup Ctenopoda. I do, however, find it difficult to use the reduction of this wayof beating resulting in such different morphologies as seen in the Gymnomera andin the Anomopoda as a synapomorphy for these two groups. In the predominantlypredatory Gymnomera, with stenopodous instead of phyllopodous legs, the meta-chronal beat is absent. In the Anomopoda the legs are very different from theserially similar legs in the Ctenopoda, but traces of the metachronal beat are presentin some groups (Fryer, 1991; also mentioned by Martin & Cash-Clark, 1995).Instead, loss of the metachronal beat could be viewed as an autapomorphy for theGymnomera. The fourth character, ‘reduction in dependence of filtration’, is truefor the Gymnonera but apparently not for the Anomopoda.

The theory of Martin & Cash-Clark (1995) could possibly be supported by thesimilarity in the antennal rami of the Onychopoda and Anomopoda. In theOnychopoda and in many species of the Anomopoda (presumably plesiomorphic)there are three segments in one ramus (endopod) and four segments, with the mostproximal often being shorter, in the other ramus (exopod), and only few natatorysetae along the sides of the rami (character 17, Figs 5, 9, 10). However, if theGymnomera (Onychopoda and Haplopoda) is still accepted as monophyletic, asadvocated earlier, then a secondary segment addition has to be accepted in theHaplopoda (Leptodora), where there are four segments in both antennal rami (Fig.3A). In accordance with this, larvae of Leptodora have only 4 and 3 segments in theantennal rami (Fig. 3B) which suggests that the extra segment in one of the ramicould have been achieved secondarily.

J. OLESEN530

The anomopodan families Daphniidae, Macrothricidae, Chydoridae, Bosminidae: monophyly andrelationships

The monophyly of the Daphniidae (including the Moininae) is convincinglysupported. Whether the Moininae would be justified as a family or should be asubfamily of the Daphniidae is subjective. This derivation of the Moininae impliesthe acceptance of reversal of the vacuolar maturation type of spermatids to thelumenal type in the Moininae. Such ‘to-and-fro’ evolution of a complex structuresuch as the vacuolar maturation type was regarded as improbable by Wingstrand(1978). As Wingstrand was aware, some other characters conflict with this idea.Both Goulden (1968) and Fryer (1991) unite the Moininae with the rest of theDaphniidae, as is also the case here (Figs 13–16). Goulden (1968) did so because of“very enlarged filter plates of the 3rd and 4th trunk limbs” (character 33, Fig. 10)and Fryer (1991) included a “specialisation of the 5th trunk limb for closing theposterior interlimb space during filtration” (character 34, Fig. 10a). No matter whichphylogenetic hypothesis is chosen, apparently unacceptable convergencies or reversalsmust be accepted. If, for example, the vacuolar spermatogenesis is assumed only tohave developed once and not have disappeared after that, as Wingstrand (1978)thought probable, then, along with other problems, the two above-mentionedcharacters (33 and 34) would have to be assumed to have developed independentlyin the Moininae and the rest of the Daphniidae. Instead it appears more likely thatthe vacuolar maturation of the spermatids has only appeared once but laterdisappeared within the Moininae in connection with the appearance of the verylarge and strange spermatozoa found in all species of Moina examined hitherto. AsWingstrand (1978) emphasizes, information on one or more of the species of Moinawith more normal small and spherical spermatozoa would be helpful. If vacuolarspermatogenesis was found in one of those species the problem of placing theMoininae would be solved, since the lack of this spermatogenesis type in otherspecies of the Moininae would then be clearly secondary. Together with the reversalof the vacuolar maturation to the lumenal type, I suggest that there have been tworeversals from the vacuolar type to the cystic type as well (in Streblocerus and Ilyocryptus),but this is weakly supported in the analysis.

The Macrothricidae, as mentioned earlier, are not supported as no potential syn-apomorphies for the group could be located. Instead, the family is indicated to beparaphyletic. The actual suggested phylogeny of the macrothricid taxa should not beconsidered as definitive since the character support is rather unconvincing (con-vergences or character reversals). Ophryoxus, for instance, is placed on the Chydoridaeclade on the basis of a dilated posterior gut section that later disappears (character 43).The main indication of this analysis is that the Macrothricidae are probably notmonophyletic and further characters to resolve their phylogeny are yet to be identified.This was apparently also the view of Wingstrand (1978), who on the basis of twospermatological characters (vacuolar spermatogenesis, character53; heliozoa-like sper-matozoa, character, 54), found only in subgroups of the Macrothricidae togetherwith the Chydoridae and the Daphniidae, indicated that the Macrothricidae wereparaphyletic. However, the result of the present analysis suggests that the cystic mat-uration type of spermatids, found in the macrothricid genera Streblocerus and Ilyocryptus,are reversals instead of retained plesiomorphies. This interpretation may be incorrectbut if so, it would only affect the position of the genera within the Anomopoda cladeand not the presumed paraphyly of the Macrothricidae.


Considering the apparent information in the spermatozoa of the Macrothricidae,it would be informative to have spermatozoa investigated in more species of thisgroup than the four examined by Wingstrand (1978).

The Chydoridae are supported by only two characters (neck organ with lateralpores, character 9; three segments in both antennal rami, character, 17, Fig. 12).Within the Chydoridae the subfamily Chydorinae was convincingly supported by asuite of characters, while no support was found for the Aloninae, which apparentlyhave been defined on the basis of a combination of plesiomorphic and insufficientlyexamined characters (e.g. head pores). Frey (1959, 1962, 1967) divided—on basisof mandibular articulation and head pore morphology (neck organ)—the Chydoridaeinto four subfamilies. A remaining problem is that the ‘not-on-apodeme articulation’type of the mandible, found in the Aloninae, should, most likely, be considered asthe plesiomorphic type and therefore cannot be used to justify the subfamily (seecharacter 20 in character discussion). A similar problem is connected with theclassical use of the so-called ‘head pores’ of the family. While having been usefulfor species identification, Olesen’s (1996) SEM data now make phylogenetic con-siderations easier. This study, among other things, established the homology betweenthe ‘head pores’ of the Chydoridae and the neck organ/dorsal organ of conch-ostracans, which is an important basis in the interpretation of the character. I havebeen unable to formulate a potential synapomorphy related to the head pores/neckorgan for the Aloninae while the Chydorinae is supported by a very constant neckorgan morphology found in almost all species.

The neck organ morphology has in this analysis been treated as two characters,a binary character that relates to the presence of lateral pores (character 9) and anordered multistate character (the only ordered character in the analysis) that relatesto the shape and general morphology of the neck organ (character 8). As a resultof using the ordered multistate character, the presented phylogeny of the Chydoridaecorresponds with what one intuitively would expect from looking at the differenttypes of neck organs (Fig. 18). It reflects—since I did not locate any contradictingcharacters—what appears to be the assumed simplest pathway for the charactertransformation in the neck organ (see Character discussion, above). The analysissuggested that the appearance of lateral pores in the neck organ should be treatedas a synapomorphy for the Chydoridae and secondarily lost in the Chydorinae andMonospilus. The only earlier published cladogram-like phylogeny that could belocated for the family is that of Smirnov (1971). Comparison is, however, difficult,since no characters are indicated on the Smirnov figure. The basic difference is thelack of support for the Aloninae in the present phylogeny while Smirnov (1971)identifies this as a monophyletic group. In general the relative position of the generaof the Chydoridae is more or less the same in the two cladograms.

The phylogeny within the Chydoridae is far from resolved with this analysis sinceonly 17 of the 32 described genera have been included. It has—despite the relativelarge literature about also this cladoceran group—been surprisingly difficult to locatecharacters of phylogenetic importance.

CONCLUSIONS

The following concluding remarks and classificatory recommendations are basedon the strict consensus tree of 218 equally short trees (Fig. 13).

J. OLESEN532

(1) The strict consensus tree supports several taxonomic groups: Diplostraca(=Onychura), Cladocera, Gymnomera, Anomopoda, Onychopoda, Sididae,Daphniidae (with the Moininae included), Moininae, Scapholeberinae, Chydoridae,Chydorinae; these can, if the result of this analysis is accepted, be used as validtaxonomic units. A suggested taxonomic hierachy, based on Figure 13, is shown inFig. 19.

(2) It is suggested that Fryer (1991, 1995) is correct in not accepting the Moinidaeas a family and that the moinids be included in the Daphniidae as a subfamily (Figs13, 19).

(3) The Conchostraca and Spinicaudata are unsupported in the strict consensustree (Fig. 13); the status of these groups is uncertain.

(4) Part of the Ctenopoda (Sididae) is supported, but unsupported with respect toHolopedium. However, Holopedium was not derived with any other of the anomopodmain groups either (Figs 13–16).

(5) Two groups—the Macrothricidae and Aloninae—are indicated to be par-aphyletic (Figs 13–15). Taxonomic rearrangements are best delayed until the presentresults find further support.

(6) The relationships between the Gymnomera (Onychopoda and Haplopoda),Anomopoda and Ctenopoda (if monophyletic) could not—due to lack of char-acters—be resolved (Figs 13–16).

NOTE ADDED IN PROOF

During the publication process of this paper a new discovery was made whichmost likely will change one of the conclusions reached. Embryological studies onvarious cladoceran taxa showed a remarkable similarity in the development of theantennule in Polyphemus pediculus (Onychopoda) and certain anomopods, not sharedby the Haplopoda and the Ctenopoda. This challenges the monophyly of theGymnomera, which were supported in this account. The subject will be dealt within a later publication.

The validity of the interpretation of the diplostracan carapace as a ‘secondaryshield’ has recently been questioned by Fryer (1996).

ACKNOWLEDGEMENTS

I am most grateful to Dr Jens T. Høeg (Zoological Institute (ZI), Copenhagen)for following all stages of this work with continuous interest. Drs Niel L. Bruce(Zoological Museum, Copenhagen (ZMUC) and Peter Gram Jensen (ZI) are boththanked for discussions and for detailed comments on earlier versions of themanuscript. I thank Drs Henrik Glenner (ZI), Jens Mossin (ZI) and Ulrik Røen(ZMUC) for useful discussions during different phases of this project. In addition, Iam grateful to Drs Joel W. Martin (Natural History Museum of Los Angeles County)and George D. F. Wilson (Australian Museum) for useful comments on a late versionof the manuscript and to Dr Geoffrey Fryer (Ambleside, Cumbria, England) forcomments on an early version of the character discussion. Dr Mary Petersen (ZMUC)is thanked for correction of the language, Dr Ewald W. Roessler (Universidad de


los Andes, Bogota, Colombia) is thanked for help in collecting specimens of Cyclestheriahislopi, and photographer Geert Brovad (ZMUC) is thanked for assistance with theproduction of the plates. This study has been partly supported by the followinggrants: Hemmingsens Legat, Emil Herborgs Legat, and Clements Legat (collectionof material in Colombia).

REFERENCES

Barnard MA, 1929. Contributions to the crustacean fauna of South Africa. 10. Revision of the SouthAfrican Branchiopoda (Phyllopoda). Annals of the South African Museum 29: 181–272.

Botnariuc N. 1947. Contributions a la connaissance des phyllopodes Conchostraces de Roumanie.Notationes Biologicae 5(1–3): 68–159.

Botnariuc N, Bayes Vina N. 1977. Contributions a la connaissance de la biologie de Cyclestheriahislopi (Baird), (Conchostraca: Crustacea) de Cuba. In Orghidan T, Jiminez AN, eds. Resultat desExpeditions Biospeologiques Cubano-Roumaines a Cuba, Vol. 2, Bucharest: Editura Republicii Socialiste Romania,257–262.

Calman WT. 1909. Crustacea. In: Lankester ER, ed. A treatise on Zoology, Part 7 Appendiculata Fasc. 3.London: Adam and Charles Black, 1–346.

Claus C. 1876. Untersuchungen zur Erforschung der genealogischen Grundlage der Crustaceen-systems. Wien: CarlGerold’s Sohn.

Dumont HJ, Pensaert J. 1983. A revision of the Scapholeberinae (Crustacea: Cladocera). Hydrobiologia100: 3–45.

Eriksson S. 1934. Studien uber die Fangapparate der Branchiopoden nebst einigen phylogenetischenBemerkungen. Zoologiske bidrag fran Uppsala 15: 23–287.

Flossner D. 1972. Krebstiere, Crustacea. Kiemen und Blattfusser, Branchiopoda, Fischlause, Branch-iura. Die Tierwelt Deutschlands 60: 1–501.

Frey DG. 1959. The taxonomic and phylogenetic significance of the head pores of the Chydoridae(Cladocera). International Revue der gesamten Hydrobiologie 44: 27–50.

Frey DG. 1962. Supplements to: The taxonomic and phylogenetic significance of the head pores ofthe Chydoridae (Cladocera). International Revue der gesamten Hydrobiologie 47(4): 603–609.

Fryer G. 1963. The functional morphology and feeding mechanism of the chydorid cladoceranEurycercus lamellatus (O. F. Muller). Transactions of the Royal Society of Edinburgh 65: 335–381.

Frey DG. 1967. Phylogenetic relationship in the family Chydoridae (Cladocera). Marine BiologicalAssociation, India, Symposium on Crustacea. 1: 29–37.

Fryer G. 1968. Evolution of adaptive radiation in the Chydoridae (Crustacea: Cladocera): a study incomparative functional morphology and ecology. Philosophical Transactions of the Royal Society of London,B 254: 221–385.

Fryer G. 1969. Tubular and glandular organs in Cladocera Chydoridae. Zoological Journal of the LinneanSociety 48: 1–8.

Fryer G. 1970. Defaecation in some macrothricid and chydorid cladocerans and some problems ofwater intake and digestion in the Anomopoda. Zoological Journal of the Linnean Society 49: 255–269.

Frey DG. 1971. Worldwide distribution and ecology of Eurycercus and Saycia (Cladocera). Limnology andOceanography 16: 254–308.

Fryer G. 1972. Observations on the ephippia of certain macrothricid cladocerans. Zoological Journalof the Linnean Society 51: 79–96.

Fryer G. 1974. Evolution and adaptive radiation in the Macrothricidae (Crustacea: Cladocera): astudy in comparative functional morphology and ecology. Philosophical Transactions of the Royal Societyof London, B 269: 137–274.

Fryer G. 1987a. A new classfication of the branchiopod Crustacea. Zoological Journal of the LinneanSociety 91(4): 357–383.

Fryer G. 1987b. Morphology and the classification of the so-called Cladocera. Hydrobiologia 145:19–28.

Fryer G. 1988. Studies on the functional morphology and biology of the Notostraca (Crustacea:Branchipoda). Philosophical Transactions of the Royal Society of London, B 321: 27–124.

Fryer G. 1991. Functional morphology and the adaptive radiation of the Daphniidae (Branchiopoda:Anomopoda). Philosophical Transactions of the Royal Society of London, B 331(1259): 1–99.

J. OLESEN534

Fryer G. 1995. Phylogeny and adaptive radiation within the Anomopoda: a preliminary exploration.Hydrobiologia 307: 57–68.

Fryer G. 1996. The carapace of the branchiopod Crustacea. Philosophical Transactions of the Royal Societyof London, B. 351: 1703–1712.

Gerstaecker 1866–1879. In Bronn’s Thier-Reichs 5(5). (No exact year is available for the establishmentof the Diplostraca, see Fryer, 1987a).

Goulden CE. 1968. The systematics and evolution of the Moinidae. Transactions of the AmericanPhilosophical Society 58(6): 1–101.

Korovchinsky NM. 1992. Sididae and Holopediidae (Crustacea: Daphniiformes). Guides to Identification of theMicroinvertebrates of the Continental Waters of the World. 3. The Hague: SBP Academic Publishing bv.

Lilljeborg W. 1901. Cladocera Sueciae, oder Beitrage zur kenntnis der in Schweden lebendenKrebsthiere von der Ordnung der Branchiopoden und der Unterordnung der Cladoceren. NovaActa regiae Societatis Scientiarum Upsaliensis 19: 1–701.

Linder F. 1945. Affinities within the Branchiopoda with notes on some dubious fossils. Arkiv for Zoologi37A(4): 1–28.

Maddison WP, Maddison DR. 1993. MacClade, analysis of phylogeny and character evolution, ver. 3.03.[Software included] Sunderland, Massachussets: Sinauer Associates Inc.

Makrushin AV, 1985. Peculiarities in the reproductive system in Polyphemoidea, Penilia avirostris andMoinidae (Crustacea, Cladocera) related to the loss of yolk by subitanic eggs. Zool. Zh. 64: 769–771(In Russian).

Martin JW. 1992. Chapter 3: Branchiopoda. In: Harrison FW, Humes AG, eds. Microscopic Anatomyof Invertebrates, Vol. 9, New York: Wiley-Liss.

Martin JW, Belk D. 1988. Review of the clam shrimp family Lynceidae Stebbing, 1902 (Branchiopoda:Conchostraca) in the Americas. Journal of Crustacean Biology 8(3): 451–482.

Martin JW, Cash-Clark C. 1995. The external morphology of the onychopod ‘cladoceran’ genusBythotrephes (Crustacea, Branchiopoda, Onychopoda, Cercopagididae), with notes on the morphologyand phylogeny of the order Onychopoda. Zoologica Scripta 24(1): 61–90.

Martin JW, Felgenhauer BE, Abele LG. 1986. Redescription of the clam shrimp Lynceus gracilicornis(Packard) (Branchiopoda, Conchostraca, Lynceidae) from Florida, with notes on its biology. ZoologicaScripta 15(3): 221–232.

Martin JW, Laverack MS. 1992. On the distribution of the crustacean dorsal organ. Acta Zoologica73(5): 357–368.

Mordukhai-Boltovskoi PD. 1968. On the taxonomy of the Polyphemidae. Crustaceana 14: 197–209.Olesen J. 1996. External and phylogenetic significance of the dorsal/neck organ in the Conchostraca

and the head pores of the cladoceran family Chydoridae (Crustacea, Branchiopoda). Hydrobiologia330: 213–226.

Olesen J, Martin JW, Roessler EW. 1997. Description of external morphology of the male ofCyclestheria hislopi (Baird, 1859) (Crustacea, Branchiopoda, Spinicaudata), with comparison of maleclaspers among the Conchostraca and Cladocera and its bearing on phylogeny of the ‘bivalved’Branchiopoda. Zoologica scripta 25(4): 291–316.

Roessler EW. 1995. Review of Colombian Conchostraca (Crustacea)—ecological aspects and lifecycles—family Cyclestheriidae. Hydrobiologia 298: 113–124.

Roessler EW, Sanchez M. 1986. Estudios sobre los ‘Entomostraceos’ de Colombia. I. Contribucionesal concocimiento de la historia natural de Cyclestheria hislopi (Baird, 1859), (Arthropoda, Crustacea,Conchostraca), con enfasis en aspectos bioecologicos y del ciclo vital. Caldasia 14(68–70): 679–707.

Røen U. 1994. Drawings of Danish Cladocera by P.E. Muller, Steenstrupia 20(5): 101–164.Sanders HL. 1963. The Cephalocarida. Functional morphology, larval development, comparative

external anatomy. Memoirs of the Connecticut Academy of Arts and Sciences 15: 1–80.Sars GO. 1865. Norges ferskvandskrebsdyr. Første afsnit. Branchiopoda I. Cladocera Ctenopoda (Fam. Sididae &

Holopedidae). Christania: Brogger & Christie.Sars GO. 1867. Histoire naturelle des crustaces d’eau douce de Norvege. Les Malacostraces. 1–145. Christania:

Johnsen.Sars GO. 1874. Om en dimorph udvikling samt generationsvexel hos Leptodora. Forhandlinger i

Videnskabsselskabet i Kristania 1–15, pl. 1.Sars GO. 1887. On Cyclestheria hislopi (Baird), a new generic type of bivalve Phyllopoda; raised from

dried Australian mud. Forhandlinger i Videnskabsselskabet i Kristania 1: 223–239.Sars GO. 1888. Additional notes on Australian Cladocera, raised from dried mud. Forhandlinger i

Videnskabsselskabet i Kristania 7: 1–74, pls. 1–6.


Sars GO. 1890. Oversigt af Norges crustaceer med foreløbige Bemærkninger over de nye eller mindrebekjendte Arter. II (Branchiopoda-Ostracoda-Cirripedia). Forhandlinger i Videnskabsselskabet i Kristania1: 1–80.

Sars GO. 1895. Descriptions of some Australian Phyllopoda. Archiv for Mathematik og Naturvidenskab, B17(7): 1–51, pls. 1–8.

Sars GO. 1896a. Fauna Norvegiae. I. Description of the Norwegian species at present known belonging to thesuborders Phyllocarida and Phyllopoda. 1–140. Christania: Aschehoug.

Sars GO. 1896b. On some West-Australian Entomostraca raised from dried sand. Archiv for Mathematikog Naturvidenskab, B 19(1): 1–35, pls. 1–4.

Sars GO. 1898. Description of two additional South African Phyllopoda. Archiv for Mathematik ogNaturvidenskab, B 20(6): 1–23, pls. 1–3.

Sars GO. 1899. Additional notes on South-African Phyllopoda. Archiv for Mathematik og Naturvidenskab,21(4): 1–29, pls. 1–3.

Sars GO. 1901. Contributions to the knowledge of the fresh-water Entomostraca of South America,as shown by artificial hatching from dried material. Archiv for Mathemetik og Naturvidenskab B 23(3):1–102, pls. 1–12.

Sars GO. 1902. On the Polyphemidae of the Caspian Sea. Annuaire du Musee Zoologique de l’AcademieImperiale des Sciences de St. Petersbourg 7: 31–54, pls. 1–4.

Sars GO. 1904. On a remarkable new chydorid Saycia orbicularis G.O. Sars from Victoria, SouthAustralia. Archiv for Mathematik og Naturvidenskab, B 26(8): 1–15, 1 pl.

Sars GO. 1993. On the freshwater crustaceans occurring in the vicinity of Christania. University Bergen Press(translation of an 1861 handwritten, University of Oslo Library).

Schminke HK. 1981. Adaptation of Bathynellacea (Crustacea, Syncarida) to life in the Interstitial(‘Zoea Theory’). International Revue der gesamten Hybrobiologie 66(4): 575–637.

Schram FR. 1986. Crustacea. New York: Oxford University Press.Scourfield DJ. 1896. Entomostraca and the surface-film of water. Zoological Journal of the Linnean Society

25: 1–19.Scourfield DJ. 1926. On a new type of crustacean from the Old Red Sandstone (Rhynie Chert Bed,

Aberdeenshire) – Lepidocaris rhyniensis gen. et sp. nov. Philosophical Transactions of the Royal Society ofLondon, B 214: 153–187.

Scourfield DJ. 1940. Two new and nearly complete specimens of young stages of the Devonian fossilcrustacean Lepidocaris rhyniensis. Zoological Journal of the Linnean Society 152: 290–298.

Smirnov NN. 1966. The taxonomic significance of the trunk limbs of the Chydoridae (Cladocera).Hydrobiologia 27: 337–343.

Smirnov NN. 1971. Chydoridae. Fauna of the U.S.S.R. 1(2). Leningrad. Translated from Russian. IsraelProgram for Scientific Translations. Jerusalem, 1974.

Smirnov NN. 1992. The Macrothricidae of the World. Guides to Identification of the Microinvertebrates of theContinental Waters of the World. 1. The Hague: SPB Academic Publishing bv.

Smirnov NN, Timms BV. 1983 A revision of the Australian Cladocera (Crustacea). Records of theAustralian Museum Supplement 1: 1–132.

Swofford DL, 1993. PAUP: phylogenetic analysis using parsimony, Version 3.1.1. Computer programdistributed by Illinois State Natural History Survey, Champaign, Illinois.

Walossek D. 1993. The Upper Cambrian Rehbachiella and the phylogeny of Branchiopoda andCrustacea. Fossils and Strata 32: 1–202.

Walossek D. 1995. The Upper Cambrian Rehbachiella, its larval development, morphology andsignificance for the phylogeny of Branchiopoda and Crustacea. Hydrobiologia 298: 1–13.

Weismann A. 1877. Beitrage zur Naturgeschichte der Daphnoiden. Zeitschr. wiss. Zool. II. UberEibildung bei den Daphnoiden.

Wesenberg-Lund C. 1952. De danske søers og dammes dyriske plankton. København: Ejnar Munksgaard.Wingstrand KG. 1978. Comparative spermatology of the Crustacea Entomostraca. I. Subclass

Branchiopoda. Det Kongelige Danske Videnskabernes Selskab. Biologiske Skrifter 22(1): 1–67.

J. OLESEN536

APP

EN

DIX

Dat

am

atri

x.W

ith47

ingr

oup

taxa

(Dip

lost

raca

)an

dtw

oou

tgro

upta

xa(A

nost

raca

and

Not

ostr

aca)

and

with

56ch

arac

ters

.‘?

’in

dica

tes

data

‘−’

indi

cate

sw

here

the

char

acte

ris

inap

plic

able

.

Documents

A phylogenetic analysis of the Conchostraca and Cladocera