phylogenetic and phenetic systematics of the …

195

PHYLOGENETIC AND PHENETIC SYSTEMATICS OF THEOPISTHOP0ROUS OLIGOCHAETA (ANNELIDA: CLITELLATA)

B.G.M. JaniesonDepartnent of Zoology

University of QueenslandBrisbane, Austral ia 4067

Received September 20, L977

ABSTMCT: The nethods of Hennig for deducing phylogeny have been adaptedfor computer and a phylogran has been constructed together with a stereo-phylogran uti l izing principle coordinates, for alL farni l ies of opisthopor-ous ol igochaetes, that is, the Oligochaeta with the exception of theLunbriculida and Tubificina. A phenogran based on the sane attributesconpares unfavourably with the phyLogralns in establishing an acceptableclassif ication., Hennigrs principle that sister-groups be given equalrank has not been followed for every group to avoid elevation of themore plesionorph, basal cLades to inacceptabl.y high ranks, the 0l igochaetabeing retained as a Subclass of the class Clitel lata. Three orders arerecognized: the LumbricuLida and Tubificida, which were not conputed andthe affinities of which require further investigation, and the Haplotaxida,computed. The Order Haplotaxida corresponds preciseLy with the SuborderOpisthopora of Michaelsen or the Sectio Diplotesticulata of Yanaguchi.Four suborders of the Haplotaxida are recognized, the Haplotaxina,Alluroidina, Monil.igastrina and Lunbricina. The Haplotaxina and Monili-gastr ina retain each a single superfanily and fanily. The Alluroidinacontains the superfamiJ.y All"uroidoidea with the fanilies Alluroididaeand Syngenodri l idae. The Lurnbricina consists of f ive superfaniLies. Ofthese the Biwadril-oidea and Sparganophiloidea superfan. nov. contain eacha single genus and fanily. The Almoidea superfam. nov. contains the fam-i l ies Alnidae and Criodri l idae and, provisional. ly, the Lutodri l" idae.The Lumbricoidea is constituted by the Lunbricidae (redefined to includethe DiporodriLidae as a subfanily), Konarekionidae, Kynotidae, Glossos-colecidae s.str ict., Microchaetidae and Hor:nogastridae (redefined to includethe Ailoscolecidae as a subfaniLy, with the Hormogastrinae and Vignysinaefornerly included). The Megascolecoidea, as previousl.y, contains theMegascolecidae, Eudrilidae and Ocnerodrilidae, although the phylogram in-dicates that the Ocnerodrilidae should be placed in a separate and prin-itive superfamily of the Lurnbricina.

*t(

Introduction

A higher, suprafani l ia l , c lassi f icat ion of the 0l igochaeta has beenattempted by Benham (1890), Beddard (1895, 1901), MichaeLsen (1921' 1928'L929, 1930), Meyer (1929) and Yarnaguchi (1953) and Brinkhurst and Janieson(1971). Of these, those of Michaelsen (1950), of Yamaguchi and of thelatter two authors are the nost pertinent to a contenporary discussionof the phylogeny of the earthworms. A study of the phylogeny of earth-worns cannot be divorced fron consideration of that of the Haplotaxidaeand the Alluroididae, the total conplex comprising the Opisthopora ofMichaelsen, (1930), the Subsect io Diplotest iculata of Yanaguchi (1953),or the order Moniligastrida and the order Haplotaxida, ignoring the sub-order iubificina, in the systen of Brinkhurst and Janieson. Janieson(L976) has reduced the orders of the Oligochaeta to two, the Lumbriculida

1/r*

Evolut ionary Theory 32L95-233 (I" larch' 1978)The editors thank C. J. Goodnight and another referee for help in evaluat ingthis paper.

O1978, by The Universi ty of Chicago for the author

r-96 B.G.M. JAMIESON

and the Haplotaxida, by inclusion of the Moniligastridae as a suborderMoniligastrina in the Haplotaxida following reinterpretation of the natureof the moniligastrid testis-sac and construction of a phytogeny of theMoniligastridae according to the principLes of Hennig.

Gates (e.g. L972) recognized no categories between the order Oligo-chaeta and the fanilial ranks, and therefore attenpted no groupings(other than an alphabetic listing) of the fanilies Acanthodrilidae, Allur-oididae, Criodri l idae, Eudri l idae, Glossoscolecidae (s.str ict.) Haplotaxidae,Hornogastridae, Lunbricidae, Megascolecidae, Microchaetidre,, s. Gates,Moniligastridae, Ocnerodrilidae, Octochaetidae and Sparganophilidae. InL976, however, Gates recognized a suprafanilial grouping, the Lumbricoidea,more restricted in constitution than that of Brinkhurst, and Janieson, toinclude those fanilies known or supposed to have ovaries with a single eggstring, the Lr-unbricidae, Biwadrilidae, Konarekionidae, Lutodrilidae, Spar-ganophil idae, Hornogastridae, Ailoscolecidae and Diporodri l idae, alL holarc.tic groups. IncLusion of the Biwadrilidae went contrary to a statenent ofNagase and Nonura (L937:348) for the tfpe-species that the ovaryrrbeconesdivided into lobesrr and the author is unable to trace the evidence forregarding the ovaries of the Ail.oscolecidae and Diporodril.idae as single-str inged. The renaining famil ies, with the possible exception of theHornogastridae, are weLl known to possess single-str inged ovaries, however.The only monothetic character of the Lurnbricoidea s. Gates is the ovariaf,one and the present study attenpts to elucidate the significance of thisshared character.

one of the rnajor differences between the systens proposed by Jarnieson(1971d, L976a) and the cLassif ication, al.beit soLely at a famil ial Level.,of Gates lies in his treatnent of those fanilial or subfamiLial taxa tlpifiedby GlossoscoLen, Mienoeltaekn, and ALma. Janieson (1971d) has argued thatAl,nn and other genera grouped with it should be renoved fron the Microchae-t idae, in which, on the basis largeLy of posttesticular spernathecae, theywere placed by Michaelsen and Gates, and that the African genera Mtcroehae-tzn and IrLtogenia have their closest rel.atives in the S. ftnerican Glossos-colecidae such as Glossoseoles and PontoeeoLes. The S. American generaDz.LLoerh,rc and GlyphidytLoerLue are regarded as belonging with ALma in theAlrninae (Janieson, 1971d) or ALnidae (Janieson, 1976a). CyLodtilus isgrouped with ALmq though as a separate tribe or subfanily but a special.affinity between Criodrilrc and the Lunbricidae is recognized (Janieson,1971d). Omodeo (1956) anticipated the author in recognizing that ALmaand Dtiloeyius were widely separate from MLeroeVwetrc and frorn the Glossos-colecid earthworms, although the group to which they belonged was not named.The family Microchaetidae was ?econstituted to contain CalLidy,i,Lus andGLyphidriLw" which the author considers closely reJ.ated to ALma and Dy,LLoe-ritso in addition to Mi.erocVnetus and TrttogerrLa. Kyrntus was also incLudedin the Microchaetidae in which Horrnogaster formed a subfanily, the Hornogas-trinae, Lwnbriew was considered by Onodeo to be descended fron a formsirnilar to CrtodyiLtn, as supported by Jamieson (1971d).

A further najor area of debate has been the classification of the FanilyMegascolecidae in sensu lato of Stephenson (1930). The alternative class-i f icat ions of Stephenson (1930), Michaelsen (1933), Pickford (1937),Omodeo (1958), Lee (1959), Gates, (1959) and Sims (1966, L967) have beenextensively reviewed by Sins (1966) and Janieson (1971b) and it wil l suf-fice here to note the chief areas of disagreenent, in the proposals of Gates(1959), Sims (1967) and Janieson (1971a). Gates (1959) recognized fourdist inct farni l ies for the Megascolecidae s. Stephenson: the Megascolecidae(racemose prostates), the AcanthodriLidae (with tubular prostates andneronephridia) and the OcnerodriLidae (prostates and nephridia as the

L97

I

; i ;b.c ' ,rorgc.Nsr| f .dOEH\,

'dqrd,rpol t rN(J rd c! . .

-C O o o tr l ' ) <t td o ( !

\o 6 (J uN o .d o F: O N \ tor H | . l dO O ' - .dr .d E 5 bO d I rA O <> d d Nco € <t h t ror aJ rd = . .d rd t r @ d o N

I (d r F\ Od td F) d (S . . S I O \O fr !6u. l I<t (J I N E t i n dD u) t ! \O o . . F- +).d I glO O | <. . CA Cr. . 3! .drA c) . .Ln 5 S N OdaO H6 r r4 t4 Aa r : i F- \Od d i - t 5 dor N

. r l N .$ . dor o ( ! . ' u) cr l ' . . , or rh < t <t: . : !A . r N ADa O.dot 'd f rn <t d 6t . . . d F- +J . .

< ' \ t I F- | .6@ +, +rcl t H dN O N .h. . N^r i F- \O , . l t r E EH d DO | ^ ts O 55 FFU)F N orr N $r4 d Z td i t r .Ot e C . . ,<F Ot ( ,_.Ol i F dd .N .=d . , Ot I C O C d O O . O F- \OOt . , { O

- N i d to. . o>iD o . ' ! ' o o. c o ch .oH 'o.xa q F c t+N h 6. . o 14 . ' .d O C O .d d cd C c dI E or o ^ H=f (D .d E 6 N E . i O o E o | ^ H od

\O O i O o Or .d i O .E€ r d E o O d orA C |J d: g l O o . 'd E\o( i C "o Or n ( !c) 'd b c) OOO Oro=

<.. o . !d {J o d r -C O o O H F).d E O sf .dH o @ (!

. \o. . ! o E (6 N -

F:orO o t r ro 'o E 3! t r . . I E o $rr l f rg j r E E d (J Or O n,d O > E . , - ( ! F: Cta O d. , . i I O.Or(! .d i ) HO . .Or: t .d O..do crr ab .O\t rO rrstE rqtrPiF) @ aJ O ,v E

^ i CJ O Or . . t rO F. d cO

c . r - : .^ . : R ;3 Si* 4 :43_t : 3"r 3EE . ; $3b og'ddEn 6 \O € O dOS O@N(J HO OtE F. € .q. i io

| ( ls f , rn Gt d or. dS Od . 0O O5 " or €.F a.nc

N4oO l-- € OA . B O ,d 1. . t r i i d( ' ) 'd H dC(> t OgH d H ^ ,HF tr . .d O o 1oN o 5 H+, oor F. , . !' . 'e t r E\o o+ S) or O!Z\ E F . l .d+J .

-oH otF. t r$ I . o cg@ oo & d o| l r € r) (d h c o o t r t r o H\oH\Etr o >. o hd aorO hd.d p 'qO { , t r O O.; "

oovo o 5 c 'd oHq) - \ to o 'o5 0Eb0 0 uE tr .Fd Q P O b0 !U q) . d \.r tr Ort ! . d O O H tr O O tr d O tr .F1

So r ! .d ub0 .c l r o .dOEt{ i r " cQs ord( j .d Ono O.+r-$.d C > (D S k U tU^ O kcO O v voJ . ! v E E)r d u ' b0.Fl

o $ qJ C (! .d .d a, ! i l .o (o C ,4v . .d H , tr o0 d @ q O O r i -CoN E 'd th N +r"o E E :s.d o A tr F . is utr . o t i (0 b .d cd .d o.6* l

-F

, . l B S tr lN'd ia(6 o o o t r S ss . ! 5.$ o sJE g.oF6l jg +5oq).H oz.cb0(/) 'dtrsoo o "Ba E,o+! . rs .$so rdc(9'Fr O to iJ O R C Slz S S tr d O F{ OA Sr \ ) S vE.F\ Si N q).d }a 'Ftol

' "$: 'S '€3'8€t- EEt8.3[ tR'E. $ F.S . . .SS I PS= !r<rsFF +s a)k.n!c \J .$, i .d(9NPE.ds!;so \ )$ s " \J .o +JoB sBis 'ds ro d:$>.!ru &p+r .$r4e{r-$trQ $Q Borq} " \J o*FJ Bscotsd P, i '3E$'$ bf€;$i .B. .F.ss.5i tNP E SB.8 gs:aFtJ $o.$r i tF$ qp!

-F! ;aF\oBE.d ,a o .a - .d st l

,BE* $P*:3HHS S'3rrS-H'* :BE{ SS s, i . is . s, 3SS *P*S 3 3 " ' iS" iN'd

s ' t i lS-! Fo Ei ; :5 R.8t ,o*t5$ 8$.3t * S a5SSS x d." teto.S S t1 8 'H.I H 8: SH R: l t€ ' :8- 3 '& 3: t 'S o t :5 'P'E' t E ' i : '$. . R 5 t t - t 5 3; t 5t t . .S' ! R f l Ri [ 8;3 ' . 8[ 3 S.t d ' i t ;F* l .S E €"5 BS R.$t? R 3 tFRtRSB'S'gs; &S;f gt ; t ; gS g€s t lg

E s, I8 E., i :838'Efr dE egSeES'A8e N 5 * € 'dt : 'AEESiS€ SE,SE 8N3-X r * | | * { IHF \OF-GO Ot €) d NiJ)

i5 NNN N to M) f4!4 !A rA 14 tov) !o g <t t t t+ st r+ i+tst =t l4q)

.fi!F

0-q)

.'.1 OU'JtDrCaotEE

.tt c)

I . . g 3. .8 ' ]&| € F oo'r{ Or

€ .. t, .! (d or il | dF- E or or' O 1, H O >\O O) t \ r l boo! r ! c} H

-}z.+ st d cO .d dor F gr'Do ; f i t F. , , I l .bF d t - € !+ 6 .dt I F.dN F O dor t N q | . . H I I t A ( i . . . @ Ork | . r { OH d i <@tn "Oo.. 14 k.r{ F. @ \Ogl H C t Ol t . t ' $ .cO ON (,Pd d Ok " Or 'FF- F. O H i H ! ' ) ' r rc)Oot C O..or F E$rC 6d . . r$ | Or o r \ |o . . Frt5 d. . t rNdor << O E 'd \O' ! H O or . r H . ' SN O.. kN !6r

H r . o k . d ao\OE .d d C r t F N $ cOO >|ILF- HOIo '5 '

. O I C F- oOF-H . F O N Ot F tD.. t )H Ctr t&dt{ . i V)t t r .d 6cA O Or F- F C G . o d o| gr5 .Or ctNdo+J r E . .1 Q d t . .ot o F: g tu . . i c tdc,d{JH ra.q

iq c ' t . .+d ENO \O€d O O 'd u) a -CrF Hc- l . r \Od' t l r d HaOF) t r9.d . €d O . . t^ E r .d . td\Jor | . I Or .+Fl'

5 @ | HN F tr FF . .d rD d Or E O O EH\O trHc)sf | (aPja . rN,^ (d O Ortr E \O .d b H b0 O. iE O\OdN rt Io- : ' c o@o) ro. . r : o . .€o d N E .r {J l r -

- . ,o tN | \o lnb' l 6 @ d HN o t o b or d . .

^ / , d o . t r€ t ro @ t tNO.d O . .Or dN.q . i H .C) H F) '9 (9 E}< OS O\O..<t utXl i O . , -OH ror^ E F.tr .d d l rtDcct .H l )oo dHF d Ot O E 14 6 _O \O . . Od

' \ . . ts

N..E oo\0 ^ H d F) d O d rO C O \O q{ N

^ . C.^ ONO!. . F.Ol

d. . c \oor t r H .ol ob or o ot or J l or \o(a >iF qPFdo or\ok^ b OrH tU od . .d H o @ H O d rn O d OIDOI F. dcrtOH d O . :O N trE.. i d .d @. bO d.-Or dCo ^t{

( !+J. Or Ot!^ ^ O . A . HO.. !O OHoo $ cC(J Fd tr o 4b$ tr (6 . C C .

^. .O ( ! nb O-.

bod Or C O N(J O d rD O O .C tr O O -El- . O.E C oN ^ r \O

@ 6o.E or o . i . .or q t ) O o H o k €. !O.dO r$) Or.Ch -

H oo H..E . EOrH : , i n O crr OC@C trPAE dO'dr l o.d! o o E cdoo o > .d @ o i Od.d. .P cd€ OOlr . .dEp .ort .c rd F:o

^ rd v q E H ( ! do arro(rN5 qOc cd rJ E i r i .a t r@ O{ E H.d E qt qt E I O ^OoO d O $Co+J.d k (dF)E OdO tr . .

-+J O .sr E b . O Z,dt i t roobOO..CO E

'o 3d F: ( t ) O .J o . / iP{J rr t +5 (J v (d .d vEoxHo(!o q(0qr O € A O .( , ' ) E r / t ! , ( ) = ' l | ' ! o > ( !dv E'OF. a) ? Soca 'd o.d ro . i dv .o t< t r t r > O O aobd .k.do) u q b

o So.d Aq (o dOo0 's l B - 'n E (0 S O'"dOF.d(J sOo ,O S*tD dO(D 'd tO ' ra lS Fi to S16F' ' }<6o:Ep 'd FtDo Bs)R idA) O\t . U $ 5Y _aJ S .d +roSco I O o.S OSAS 'd $N q) + \ ) t r O'rrv F € $i $ F (a 'd:{N . t4E . t6\0 S :3 O-xa) !D .dBs 'd$B F(!€Fr 'sr \ )6 's! -B S 'dksd(J tooEF o00p$ b, &BEG$c|.c) CdP$. iJ N UrQ \JCq.\ : sO$ rv. . .dOOq OO

,s € p3s.:R'*" t ' *H6t.B P 3' ie E.:S $€PH.n5€1Et3 $$,cO & tsS _rr+r-Q{J !4q: F$r F '$q\J({$.FA D. QdS6$oO€N A.$+).J o uaaq r3utr i s N :s &Etaosci - . (J .c!d&o;E .d -s

" "oQ .SiS<tsa3.S*.S3o s 3o*l i .oS $'*r i iR "* . ' . . '&{ $*

lrotr6 +'F! .$NOFrS6$NqoQ.!N .Jq:S .+s Nr\)NtJtrq). '6OO NHd.$ cd-Q !n !a c_&rtr ! !$ s ' " Q.q Q' t ' i o I o '$( , j+<Nra s).q oc'r iq" \ ] 9$$i{ o.F . f $

Hd"BOta.F\\ BoN O.iN-Fi ' ! r . i i 's ! .d- l 'n 'p BoO I Oi ooNts Jr -EN

3S 3'8 ' iS * R3.Sg'd3 E b'$3 t b t bI3€?€5t:d 3* I .36 Se..8 3 " t 8.d ON trqi ia"u $NN O\J€ O)F rE O E O E'S6.Ft k. i rN Oq O.. O.H$ (d O € : } )SrO tr AON S< SS S) l 'dF u r !< SO,N<N<QilNPN I F{< (oN$.cO.cF.-dS'$trOOob,3N .dBScotsOSoiJ Fb'+ ' ! r bNNtrNt ' )$ OrSoof O&rrSrFqtA.N

EN * Q " j A U *F S I ;$ I I r Q E A d A d E 3 I I I X S, i E I I E SE I S 3S $,1 ?S S*i++t l l

d N vrr+tn \oF.co or s is : : 3 I I 3 I R s NR sK

l -98

B.G.M. JAMIESON

Acanthodrilidae but hearts confined to X-XI, intestinal origin in or usualLyanterior to XIV, calciferous t issue in IX-X). Sins (1966), approved Gatesclassification as a result of a taxonometric investigation but later (L967),while retaining the same groupings and the criteria for these, suggesteda special affinity with each other of three of the groups which becane thesubfanilies Ocnerodrilinae, Acanthodrilinae and Octochaetinae within thefanily Acanthodrilidae, the MegascoLecidae renaining as a distinct famil.y.In a series ofpapers Janieson (t970a, L971a' b, c, e, 7972a, b, L974a)and Janieson and Nash (1976) have refuted division on the basis of pros-tatic norphology and of hoLonephry versus neronephry, as proposed byGates, on the grounds that this would split homogeneous groups even atthe generic level. Division of the Megascolecidae into three subfanil ieswas proposed: the 0cnerodrilinae, Acanthodrilinae and the Megascolecinae.Mrile the constitution of the 0cnerodrilinae did not differ fron Gatesl0cnerodriLidae, the author (Jamieson, 197lb) presents evidence in agree-ment with Gatesr view that ocnerodriles are very distinct fron other megas-colecoids and (1971d) questions the belief that EudriLidae (also placedin the Megascolecoidea) are descended fron ocnerodriles. The constitutionand definition of the AcanthodriLinae and Megascolecinae s. Jamieson isradical ly different fron that of Gates and of Sirns. Full definit ion ofthese subfanil ies and their tr ibes is given in Jarnieson (1971a-c) but thefollowing nay be noted in relation to the phylogeny discussed bel.ow.The MegascoLecinae consists of three tribes, the Perionyctrini, theDichogastrini and the Megascolecini. The Perionychini have tubular toTacenose prostates, are partly or wholly holonephric and have the naleand prostatic pores on segment )ffIII, the negascolecin arrangenent ofpores. The Dichogastrini are neronephric with, caudally, a stonatenephridiun median to astomate micromeronephridia; prostates are tubularto racemose; and the nale pores are acanthodril.in or megascolecin. Theauthor has since formed the opinion that the Dichogastrini nay be at Leastdiphyletic. This is suspected because aLL African dichogastrins haveacanthodrilin rnaLe pores and the Acanthodrilinae are well representedin Africa while perionychins (with the rnegascolecin arrangenent) areabsent, while Australian dichogastrins have megascolecin (perionychin)pores and perionychins are very numerous in Australia. The third negas-colecin group, the MegascoLecini, typical ly has megascolecin nale pores,racemose or tubular prostates and is exclusively meronephric with verydiverse nephridial conditions, often including multiple nephrostones persegment; any nedian stomate nephridium is (always?) enteronephric. TtreAcanthodrilinae a1.e heterogeneous with regard to the nale pores andnephridia, but the pores are usually acanthodrilin (prostate porespaired in XVII and XIX, nale pores in XVIII) and never show thelnegascolecin arrangenent and the nephridia are usually holonephric,though partial or total neronephry has supervened in sone; prostatesare tubular or rarely racenose.

The choice of species for the phyl.ogenetic analysis wil l al lowdetai led investigation of the validity in tef lns of phylogenetic system-atics of former classif ications of the lumbricoid, especial ly alnid,genera but less profound evaLuation of the affinities of the megasco-lecoid genera (Table 1).

MethodologY

The rnethods here enployed are (1) a rnodification of the phylogen-etic systenatics of Hennig (1966) and (2) a phenetic nethod derived fronprinciples reviewed by Sokal and Sneath (1963) and Sneath and Sokal(!g7S), here using the set of attr ibutes selected for the phylogeneticanalysis.

L99

PHYLOGEM OF THE OLIGOCHAETA

(1) Phylogenet ic

Methods for deducing branching sequences representative of theevolution (phylogeny) of animal groups require decisions as to thedirection of change which has occurred in attributes which exist in morethan one state. In terns of nunerical, phenetic taxonony these attrib-utes may be binary (e.g. dorsal pores absent or present), ordered nult i -state (e.g. male pores in segnent L2, L3,14.. n) or d isordered nuLt istate(e.g. caudal nephrostones per side one, or nore, or none). Sneath andsokal (L973:1.0) who insist that evolut ionary branching sequences nustbe inferred largely fron phenetic relationships anong existing organismsnevertheless indulge in such decisions as to direction of change and,indeed, Sokal has contributed a valid and precise nethod (Camin and Sokal,1965) for deducing the most parsimonious phylogeny for organisms witha given set of characters. Their nethodology provides a mechanisn fordeal ing with ordered nult istate characters, that is f f t ransformation ser iest lfrom the primitive (pl,esionorph) to a sequence of advanced states (apomor-phies) in the terninology of Hennig (1966). Such al ternat ive aponorphiescannot be satisfactorily treated in Hennigrs nethod even when regarded(e.g. 8a11, 1973) as non-congruent aponorphies. This problen is especial lyevident when dealing with translocation of organs in evolution throughsuccessive netameres in netamerically segmented aninals and is furtherdiscussed, with an attenpt at solut ion, beLow. We are ini t ia l ly, however,concerned with the argunentation for deducing direction of change in allcharacters, whether binary or mult istate.

SeveraL authors have proposed sets of criteria which they considerto allow deduction as to the direction of change, which state of anattribute is prinitive (plesiornorph) and which advanced (aponorph).While the criteria proposed are in some cases inacceptable, often onthe grounds of circularity, it is important to note that decisions ofa phylogenetic nature are not necessarily less valid than deductions inother areas of science where inferences are drawn fron observed phenonena.The criticisn that they are untestable and therefore are not to 6e regard-ed as part of deductive logic seems inappropriate when it is consideredthat, where the ain is nerely the most parsirnonious phylogeny, the effic-acy of nethods such as that of Canin and Sokal in retrieving a hypothet-ical phylogeny can readily be demonstrated and that the accuracy ofinference as to direction of change can also be shown to be high on thebasis of rr inventedt ' phylogenies for hypothet ical aninals (e.g. Caminrsimaginary aninals Sokal, 1966). The di f f icul ty of convincingly proposingcriteria for.diagnosing plesiomorphy and aponorphy appears greater thanthat of intuitively recognizing these states for a group which is wellknown to the worker concerned. The tfunreconstructed logictt (sensu Hul1,1970) of intui t ive decisions may be retr ievable for each character inturn but the erection of criteria for the set of characters nay beelusive.

Select ion of characters

The characters selected were those features of the general aratomyof the oligochaetes concerned, embracing as nany organ systems as possible,which presented characters resolvable into nore than one state. Thisapproach appears to differ fron that of Hennig (1966) and Brundin (1965,1966) who commence by a rfsearch for the sister-grouprf and for the internalaponorphies which define grol4)s.

240

Table 2

Attributes EnrPloYed

Two State

1 Prostornium attached to peristolniun (p) or separate (a)

2 Lateral lines present (p) or absent (a)

5 Dorsal porcs absent (p) or present (a) .4 Caudal

'cross section not quadrangular (p) or quadrangular (a)

5 Anus terminal (P) or dorsal (a)6 Copulatory alae absent (p) or present- (a)

7 Claspers tearing rnale poies absent (p) or present (a)

8 Genilal lobes or clasplrs lateral to nale pores absent (p)

or present (a)s toni i tudinai iubercula pubertat is absent (p) or.present (a)

10 Sonit ic setae in 4 bundles (p) or perichaetine (a)

rr penial setae dist inct fronl any other genital setae absent (p)

or present (a)t2 tongitudinai ly grooved genital setae absent (p) ol Present (a)

13 Proltate-l ike'setal glands absent (p) or present (a)

t4 Cli tel lun uni layered (p) or nult i layered (a)

15 Dorsal vessel corrt inued onto the pharynx (p) or not (a)

16 Subneural vessel absent (p) or present (a)

17 Muscular thickening of iniest ine absent (p) or present (a)

18 Intestinal caeca absent (p) or Present (a)

i9 Intestinal typhlosole absent (P) or present (a)

20 Paired enteroiegnental organs absent (p) or present (a)

2t t lolonephric (p) or neronephric (a)

22 l tephri 'dial bl ldder absent (p) or present (a)

23 t leptrr idiat caeca absent (p) or present (a)

, i i" i ra"r nephridia exonephric (p) or enteronephric (a)

25 Testes in 10 (p) or absent fron 10 (a)

26 Testes in 11 (p) or absent frorn 11 (a)

,, Male pores per i ide, two (p) or one,-by fusion (a)

28 Purely muscir lar nale bursa absent (p) or present (a)

,g crtttdlrar rnale bursa absent (p) or present (a)

30 Ovaries in 12 (p) or absent from 12 (a)

31 Ovaries in 13 (P) or absent fron 13 (a)

SZ Ovaries mult istr inged (p) or single-str inged (a)

33 0varies free (P) or in septal charnbets (a)

34 Spernathecae piesent (p) or absent (a)

35 Sperrnathecae a pair oi median per segnent (p) or nultiple (a)

ta iiu*"tfto""o discrete (p) or cbmnrunicating with the oviducts (a)

Olrl Crdered i"tultistate

1 Supra-oesopha.geal vessel absent (p), oesophageal only (a1)

oi oesoPhageo-PharYngeal (a2)

2 Connissurats t trroughiut (p) or discrete hearts to 11' 12 or 13 (a1-3)

S Nunrber of segnents with ol iophageaL gi,zzatd o (p) 1, 2,3-!2 (a1-12)

+ Sugt"nt of f i rst oesophageal gizzatd (*), 5-13 (a1-9)

5 First intest inal segnent 9 (p), 10-38 (a1-29)6 Last male pores posterior to 11 or equivalent by 1 (p) 2-16 segnents

(a1- 1s)

Dl" l Disordered Mult istate

1 Seta!. interval cd = ab (p) or ) ab (a1) or< ab (a2)2 Caudal nephrostones per side one (p) or nore (a1) or none (a2)5 Testis-sais absent (p) or segnental (al) or I ' intraseptalr ' (a2) or

capping the testes (a3)4 Seminal vesicles extensive (p) or short (a1) or absent (a2)

5 Male pores anteriorly on the cl i tel lun (p) or anterior to i t (a1) oron poster ior region or behjnd i t (a2l

6 Pro!tates absent (p) or tuiru).ar (a1) or t t tbuloracenose-racemose (a2)or capsular (a3) or euprostat ic (a4)

7 Last lperrnathecae pletesticular (p) or test icular (al) or post-test icular (a2)

8 Sperrnathecae adivert iculate (p) or with senfrral divert iculun (a1) orwith non-selni.nal divert ictt lul t (a2J

9 Prostates absent or arl 'angenlcnt acanthodri l in (a1) or nicroscolecin(a2) or negascolecin (a3)

20L


The phylogenetic procedure adopted here agrees with that of Hennigand Brundin in recognizing phylogenetic affinity only on the basis ofshared advanced characters (synapomorphies) and, mlike Canin and Sokal,ignoring symplesionorphies though the nethodology of the latter workersappears no less val id.

Hennigts node of constructing phylogenies on the basis of synap-onorphies is duplicated by conputer means by sumning shared characterstates of each species (OTU) against each of the other species so as togive a natr ix of scores equivalent to a matr ix of coeff ic ients ofresemblance (here computed as distance). 0n1y i lposi t ive natchesrr(synaponorphies) are counted and the absolute sum of the natches betweentwo OTUIs is used as the coeff ic ient of s ini lar i ty without recourse todivision by the nurnber of characters. A special procedure, and netrichas been adopted for any character for which different states are differ-ences in metameric location only. An example is character 0M 6 in Table 2,Last male pores poster ior to segnent 11 (or moni l igastr id equivalent) by1 (plesiornorph state) 2-16 segments (15 aponorPh states). This charac-ter may be regarded as having 15 sequential aponolph states, beyond theplesiomorph state, bel ieved to represent an actual poster iad segnent bysegnent nigration of the pores in evoLution. To al.l.ow each of thesesegnental positions the vaLue of a fulL character would have grosslyoverweighted position of the rnal.e pores telative to other characters.A single character only is therefore used. Two species sharing the saneterminal state, whether at 1 or 15, or intermediate, score a sini lar i tyof 1 for the character. Where their states differ the sinilarity is con-puted by dividing the smaller by the larger; in this way the score forthe character cannot exceed 1. In practice the same metric can be usedfor all characters, including the two-state characters.

Computation was nade using a nodification of the rfclustrfr progran

of the University of Queensland Departnent of Conputer Science (Andrews,

tg76). The program was nodified to ensure that for disordered nultistate

characters only identical aponorph states received a score (unity), sharedplesiornorph or non-congruent aponorph states scoring zeto.

The similarities in the natrix were cLustered by the nearest neighbour(highest single l inkage) nethod (Fig. 1). Those cr i t ical of this cluster-ing nethod wi l l recognize that only i t reproduces Hennigts rnethod of es-tablishing affinities fron synapomorphies. (A corollary of this is that

i f Hennigrs nethod does val idly retr ieve at Least parsinonious phylogenies,

rejection of nearest neighbour clustering in nunerical taxonony on thegrounds that it tends to produce chaining nay be unwarranted). An ordin-ation was also prepared frorn the similarity natrix by the nethod of prin-

cipal coordinates. I t was considered of value to relate the f i rst phy-

logram to the first and second axes of the ordination, giving the threedimensional reproduct ion shown in Fig. 1. The f i rst phylogran is al ter-natively ca1led a hennigram, in deference to Herulig, while the three

dimensional representation is tenned a stereophylogran.

Determination of direction of change

The criteria for recognizing plesionorphy or aponoryhy of a char-acter suggested by Hennig (1966) have been reviewed and cr i t ic ized byCol less (1967) and Sneath and Sokal (1975). Sporne (1948), Marx and

Rabb (1970), crowson (1970), Ross (1974), Harper (1976) and Lien (1970;

202

B.G.M. JAMIESON

the latter fol lowing the work df Inger) have also suggested criteria.The criteria of Marx and Rabb (1970) have been paraphrased and rnodifiedby Ball (L974). Although Ball 's decisions as to direction of change arenot questioned, the criteria cannot be considered wholJ.y satisfactory.For example, the first criterion is that of uniqueness of a character.A decision concerning the phylogenetic relationships of the fanilies ofPaludicola having been nade, a character state unique to a derived familyis thereby inferred to be aponorph. There is, perhaps, tnore circularityin this cri terion than is desirable.

Criteria for deteflnining direction of evolutionary change

In the present study, the fol lowing criteria for diagnosing plesio-norphy or aponorphy are recognized and the characters employed are listedunder the criteria most pertinent to deternination of direction of changein each character. In keeping with previous renarks, it is consideredthat while the criteria may have unsatisfactory aspects, experiencedoLigochaetologists wil l not dispute the direction of change inferred foreach character. Construction of the most parsinonious phylogeny is centralto each criterion.

L. Rel.ative abundancea, The greater the nunber of anneLid groups (e.g. archiannelids,

polychaetes, ol igochaetes, leeches, branchiobdell ids etc.) possessinga character state (including absence), the greater is the likelihoodthat the character state is pLesiomorph.

The crux of this criterion is that it is nore parsinonious to con-sider that a character state widespread in diverse groups has been car-ried over frorn ancestral forns than to postulate independent origin ofthe character state in each groqp. Like most phylogenetic criteria itnecessitates d prLorL recognition of groupings which nust be either in-tuit ive or phenetic.Characters : TS1r3-13, L5?,L7-2L,24,29,32?,33; DM3r6r9.

b. A character which is l inited to yet widespread in ol igochaetes(i.e. which occurs in diverse rnorphologicaLLy defined groups of the Oligo-chaeta) is Likely to be plesionorph for the Oligochaeta. The presenceof spernathecae (TS34) is an exanple of such pl.esiomorphy. An exceptionis the widespread occurrence of the nultilayered condition of the clit-el lun which, i f leeches a?e taken into account, aLso transgresses crit-erion la. Application of the set of cri teria here proposed to the tot-ality of characters selected suggests that the nultilayered conditionof the cl i tel lurn is derived, as is general ly accepted by ol igochaetolo-gists though it is not beyond question. While this exception showsthat a derived condition nay be more connon than a prinitive conditionwithin a group, lack of universali ty of this cri terion is part ly attr ib-utable to the difficulty of defining what is neant by "widespread".It clearly refers to occurrence in a large proportion of divergent sub-groups rather than nunbers of species but recognition of such subgroupsa prLorL to a phylogenetic analysis creates circularity rmless initial(e.g. phenetic) classif ication of the group be permitted. Such a priortclassification seems trnavoidable and is evident in all phylogeneticanalyses of other workers which the author has studied. Where an apom-orph character evolved very early in the history of a group, as pres-unably was the case with the developnent of the multilayered clitellun,

203


it nay nevertheless be widespread in the sense defined and such connon'ness wi l l not ref lect plesionorphy for the group as a whole.Characters : TS25,26,30?,31 (opisthopores only?) i 34.

c. A character state, whether restr icted to ol . igochaetes or not,is considered aponorph if it occurs in few oligochaete taxa and if thesecannot be considered plesionorph if a parsimonious phylogeny on the basisof all characters is to be produced. This approaches the uniqueness crit-er ion of Bal l but, by invoking parsinony, avoids the subject iv i ty ofrecognizing a derived subgroup prior to determination of direction ofchange. Mere restriction of a character state to few taxa is not con-sidered a criterion of apomorphy but where, as is particularly well exem-pl i f ied by the Moni l igastr idae, restr icted character states occur ina group which could not be conceived as ancestraL to groups lacking thecharacter ist ics (or, object ively, would not give a parsinonious phylo-geny if so regarded) aponorphy is indicated.Characters : TS1r4-8,9?, 10,LL? r tz, t3?,L8,20 rzL,24,29?,32-36; DMS, 6 17?,8'9.

2. Conplexi tya. A character state which is restricted to or nainly found in

oligochaetes with the most complex organization is likely to be apom-orph. This also relates to Bal. l rs uniqueness cr i ter ion but introducesthe concept of cornplexity. Higher complexity is considered to be an in-dication of apomorphy of the total organisn and to be nore objectivelyassessable than whether a group is "derivedrr. For instance, Phenetima,with several types of nephridia, with enteronephry and exonephry, intes-t inal caeca, per ichaet ine setae and test is-sacs, is c lear ly nore conplexthan an al luroidid or a haplotaxid.

To some extent to recognize conplexity is to postulate apomorphyfor each of the characters or systens showing such conplexity. Thereis also no dist inct borderl ine between complexi ty and relat ive sinpl ic i ty.The Line between primitive and derived taxa is, however, equally arbit-Taty. This criterion is in sone regards a restatement of criterion 2c,for a complex aninal could not general"ly be regarded as basal in a par-sinonious phylogeny. The l ist of characters which fol lows is restr ictedto aninals with a high degree of conplexi ty,Characters : TS4-8, t0,12,t7,L8,L9?,33,35,36?; DM2.

b. A character state which is restr icted to or nost ly found inol igochaetes with the least conplex organizat ion is l ikely to be plesio-rnorph. The arguments given for 2a pertain to this criterion also.Characters: T52,I4,27,30; 0M2,6; DM1,4.

3. Ecological special izat ionA character state which is significantly nore abundant in oligochaetes

with a part icular ecological special izat ion is Likely to be apomorph.This is the frecological special izat ioni l cr i ter ion of Marx and Rabb andof Bal l . Sone circular i ty is involved in determining which node of l i feis special ized and which pr ini t ive. In this case a l innic rnode of l i fein earthworm-l ike forms (ahnids, Eisen'LeLla)appears secondary ( i .e.specialized) and to correlate with a quadrangular cross section and dor-sal anus. The haplotaxids, also l imnic but plesionorphic in nost char-acters and therefore ecological ly unspecial ized do not share thesecharacter states.Characters: TS4,5.

204

Table 3

Phylogeneticnearest neighbour

47. Spenceriel la48. Lanpito50. Pheretina47. Spenceriel la47. Spencer ie l la47. Spenceriel la

' PheneticR nearest neighbour

35.3 45. Digaster b.

30.9 48. Lanpito3L.0 47. Spenceriel la31.0 47. Spenceriel la30.9 47. Spenceriel la

;0.T.u.

1. Haplotaxis2. Al luroides3. Brinkhurstia4. Standeria5. Syngenodrilus6. l ,bni l igaster7. Desrnogasterr8. Biwadri lus9. Sparganophilus

10. Alna11. Glyphidri lus

12. Dri locrius a.15. Dri locrius h.14. Glyphidri locrius15. Cal l idr i lus s.16. Cal l idr i lus u.17. Microchaetus18. Glossoscolex19. Pontoscolex20. Kynotus21. Hormogaster22. tlignysa23. Hemigastrodri lus24. Diporodri lus25. Ai loscolex26. Kornarekiona27. Lutodri lus28. Criodri lus29. Lunbricus30. Eiseniel la31. 0cnerodri lus32. Eukerria33. Eudri lus34. Nenertodrilus35. Microscolex36. Diplotrena f.37. Diplotrena sP'SS.ochtochaetus39. PerionYx40. lleteroporodrilus41. Diporochaeta42. Cryptodri lus43. Dichogaster b.44. Dichogaster o '45. Digaster b.46. Megascolides

47. Spenceriel la48. Larrpito49. Celer ie l la50. Pheretina

2. Al luroides 50.0 2. Al luroides 14.34. Standeria 46.3 4. Standeria 6.84. Standeria 45'1 4. Standeria 6.43. Brinkhurst ia 45.1 3. Brinkhurst ia 6.4

19. Pontoscolex 43.5 2. Al luroides 14.87. Desnogaster 37,5 7. Desrnogaster 7.16. Moni l igaster 37.5 6. Moni l igaster 7-L

17. Microchaetus 42.5 9. Sparganophilus t7,726. Konarekiona 42.3 26. Komarekiona 12.627. Lutodri lus 42.314. Glyphidri locrius 37.? 14. Glyphidri locrius 10.412. Dr i locr ius a. 37.0 L6. Cal l idr i lus u. t3.716. Cal l idr i lus u.13. Dr i locr ius h. 56.0 13. Dr i locr ius h. 9.815. Cal l . idr i lus s. 35.0 14. Glyphidr i locr ius 7.613. Dr i locr ius h. 36.8 13. Dr i locr ius h. 7,616. Cal l idr i lus u. 34.1 16. Cal l idr i lus u. 4.215. Cal l idr i lus s. 54.1 15. Cat l idr i lus s. 4.225. Ai loscolex 36.5 25. Ai loscolex L2.619. Pontoscolex 36.1 19. Pontoscolex 14.018. Glossoscolex 56'1 18. Glossoscolex 14.018. Glossoscolex 37.4 34. Nenrertodri lus 18.925. Ai loscolex 35.1 25. Ai loscolex 11.929. Lurnbricus 38.0 23. Hernigastrodri lus 4.921. Hormogaster 37.3 22. Vignysa 4.929. Lumbricus 36.1 30. Eiseniel la 7.621. Hornogaster 35.1 21. Homogaster 11.917. Microchaetus 37.7 9. Sparganophilus L2.611. Glyphidr i lus 38.0 15. Cal l idr i lus s. L7.413. Dr i locr ius h. 37.5 13. Dr i locr ius h. t7.330. Eiseniel la 34.7 30, Eiseniel la 7.329. Lunbricus 34.7 29. Lunbricus 7.332. Eukerria 4t.2 32. Eukerria L2.835. l , l icroscolex 40.1 31. Ocnerodri lus t2.834. Nemertodri lus 35.1 34. Nenertodri lus 8.133. Eudr i lus 35.1 33. Eudr i lus 8.136. Diplotrerna f. 54.5 36. Diplotrena f. 4,937. Diplotrena sp. 54.3 55. Microscolex 4 '943. Di lhogaster b. 33.1 43. Dichogaster b. 8.143. Dichogaster b. 33.6 43. Dichogaster b. 7.L50. Pheretirna 34.3 40. Heteroporodri lus 10.542. Cryptodri lus 32.7 42. Cryptodri lus 5.249. Cel l r ie l la 35.8 35. Microscolex 8.1

40. Hetexoporodri lus 32.7 40. Heteroporodri lus 5.237. Diplot iena sp. 33.1 44. Dichogaster o ' 4.843. Di lhogaster b. 33.4 43. Dichogaster b. 4.842. crypt;dri lus 34.0 42. Cryptodri lus 7'8

8.7

3.95.97.87.6

R = Relat ionship expressed as dissirni lar i ty by deducting total synaponorphiesfor two 0,T,Urs fron total possible apomorphies (51).D = Phenetic sini lar i ty expressed as dissimilari ty by deducting fron 100 thesun of posit ive and negative matches for two o.T.Urs, divided by the numberof actual cornparisons, e4lressed as a percentage..

205


The criteria of rrmorphological special izationrr and rrgeographical

restrictionrt of Marx et al, though appearing valid, cannot readily beapplied to the ol igochaetes in this study.

(2) Phenetic

The phenetic nethod employed in this study differed fron conven-tional neo-AdansoniannumericaL taxonomy in that the attributes usedwere subjected to a prLori, seLection, that is they were not a randomsaurple of attributes though derived fron €ls many organ systems as prac-t icable. The basis for this selection was that the attr ibutes werethose used for the phyl.ogenetic systenatics, being attributes divisibleinto plesionorph and aponorph states. The rnethod was phenetic in thatnegative natches (shared plesiomorphies in Hennigian terrns) contributedequally to the sirniLarity coefficient in conparison with positive natches(shared apornorphies ) .

The program was adjusted so that onLy identicaL states of nultistateattributes, whether positive or negative, were recorded as natches, nofractional scores being aLlowed. The sinilarity natrix was again con-puted with the rClustrr program and a nearest neighbour (highest singlelinkage) dendrogran (phenogran) was generated.

ResuLts

The results of the phylogenetic anal.ysis are shown in the hennigram(Fig. 1) and stereophylogran (Fig. 2) and those of the phenetic analysisin the phenogran (Fig. 3). Lack of space does not pennit inclusion of thesimilarity natrices but the character states for all 0TUts are reproducedin Fig. 1 which shows the synapornorphic relationships and Table 3 listsphylogenetic and phenetic nearest neighbours.

Discussion

The discussion wil l centre on the phylogenetic analysis using frac-t ional scores (hennigran, Fig. 1 and stereophylogran, Fig. 2). Conparisonwil l be made with the phenetic analysis (phenogran, Fig. 3).

Although Hennigian principl.es have been applied in developing thehennigram, a rigidly Hennigian interpretation of the Hennigram will notbe nade as the author has sone reservations concerning the view of spe-ciation and phylogenesis which l ies behind Hernigrs nethodology. Hennigrsnethods have been used, with nodification, chiefly for their value infacilitating treatnent of large anounts of data (especially where con-puterized), for their relative objectivity, their repeatabil i ty and forthe heurist ic quali t ies of the resultant phylogran.

One of the chief reservations relates to the view expressed by Hennigand by his protagonist Brundin that evoLution to sister species and thencehigher groups proceeds by dichotony of single species (both the dichotonyand origin from single species are questionable) and the belief that agrouping of forns which have evolved siniLar characterist ics by paral lel-ism is not taxononically val id. The author has previously explainedvariability and apparent revetsion to prirnitive characters in a groupin terms of both intra- and inter-specific and rnorphogenetic lability

206

--&r<oo

tf-H

ddQQ

39Er

J)

QE

<=FJ

<dIQooOFe&

XEz?ooc0

-@aJ d- <3- EUJ i

6 J-A@P- tAO= /e)tsae<6urJ<doJo<QEsax

; ;8e;6s98?5FoddoQoJdoYcqouu6<=F9zz>JtslZ<-6ZZ<-JL

3_33.1;&94;.9--fzI

ffiffiGffi(.H

n

ll

I

aa

Fig. 1. Hennigram (phylograrn) for all fanilies of the opisthoporousoligochaetes constructed by nearest neighbour clustering, the nearestneighbour of an entity (oru) being that entity sharing with it thegreatest number of advanced (aponorph) character states. Plesiornorphstates for all characters are indicated by blank ovals. For each of the6 ordered nultistate and 9 disordered nultistate characters successiveapomorph states, as given in Table 2, ate indicated alphabetically byupper case letters. For ordered mult istate character 5, in which thereare more than 26 apomorph states, apomorph state 2g is indicated by alower case frcrf . Asterisks denote rnissing or inadnissible data.

207


or plasticity in ancestral populations for the characters concerned(Jamieson, L97Ld).

These ancestral rpopulationsr may have been groups of closely rel-ated species evolving similar characters in connton by virtue of sinilarresponses, to selection, of gene pools which were basical ly siniLar becauseof relationship. An example is the paral lel acquisit ion of mechanisnsfor internal fertilization in different l.ineages of the Eudrilidae ar-gued for by Janieson (1969). This view of evolution of a given groupby paral leLisn from a stock of interrel.ated species is fundanentaLly atvariance with the principles of Hennig and Brundin (cf. origin of nar-supials and eutherians from a single species, Brundin, 1966). The viewhere presented is that evoLution comnonly, if not norrnaLly, proceeds byparalLelism fron closely related species.

Reappearance of the plesiomorph condition, two pairs of ovaries,in three of the known species of rnegadriJ.es (EnantLodrLlus boreLlii,DipLoeardLa sande!,si and Glyphi,drilw kukent?wli) or of two pairs of malepores in the negascolecid genus HopLochaeteLla are probably exanples ofnorphogenetic plasticity although they nay more questionably be the resuLtof persistence of these characters in some species of a conmon ancestralpool of species. Sini l .arly, occurrence of alternative condit ions of thenale tenninalia (for instance the acanthodriLin and megascolecin con-dit ions) nay indicate ( intraspecif ic?) variabiLity in these charactersin ancestral stocks rather than the occurrence of only one or the otherin a single ancestral. species.

The possibi l i ty of rnisclassif ication, that is spurious associationor separation of clades in the hennigraq also has to be reoognized andis suspected for the 0cnerodri l idae.

Two chief sources of rnisclassif ication would seen to exist in Hen-nigian methodology: (1) fai lure to use the total i ty of extant speciesof the group under consideration, or (2)' more fundamental, the factthat determination of groupings is essentiaLly probablistic. Thus thesnall"er the nurnber of symplesiornorphies, the smaLler is the confidencethat the consequent grouping is not the result of convergent acquisitionof the shared apomorphies.

MegascoLecidn e and EudrLlidne

The Eudrilidae (represented by EudrLlus and Nemerfudri'Lrc' sub-families Eudrilinae and Pareudrilinae respectively) appeaT as the ples-iomorph sister-group of the Megascolecidae excluding the Ocnerodrilinae.Their relative pl.esiomorphy is striking in view of the tmique and highlyaponorph developnent of mechanisms for internal fertiLization in bothdivisions of the Eudri l idae. The EudriLid-Megascolecid grouping is thesuperfamily Megascolecoidea of Brinkhurst and Jamieson and retains thisstatus here but the ocnerodrilinae were also included by us in the Megas-colecoidea and wil l be discussed beLow. The sister-group of the megas-colecoid grouping in the phylograrn includes lunbricids and glossoscolecidssensu Lato but notably excludes the aLnids. This sister-group, hereterrned Ltrmbricoidea, is conposed of Lumbria,$ and EieewLella (Lunrbricidae) ,and recently erected famil ies Diporodri l idae and Ailoscolecidae (a11 fourtaxa being part of the Lurnbricoidea in the narrow sense of Gates, L976),and the core of the glossoscolecids s. lat. Brinkhurst and Jamieson,

,.9ooO b0'Ff Fl id+r t r .c (0 o ' t r EO.. . { B 9eo (D tstdtr . ' {5+rh.€."{ ^Cl O.b0

8€5.53€.$b0 O .Fi F{ C d'tJ.'{ (.) +J tr{ O

F{ t! 'F{ tri Oo xtr xo o.d-O

.F{ -O

.il .r-{ +Jo "d q.r +J5.d ht iN d o0( l )0( | ) ( l ) ( )+JH+r CO. ' - {O O O.rr OtH X

8EF$*'d.gsP .o 'qoP |n o a.(6 o' / t t r ' . Ftr-r ,g.HOgul0()+Jp{o o d EO . . r ' l 3 O-l t r- t r+J A(d.rrqi O d. r (d.HO O g A CFt€

.o.F{ O X'Fr C)q 'OXoEcO t4 Ft d d.F{

. ' - { d O t |<+{ dr i -C E O l i - l'E E F 6 * 8.P-at Gl o.d 5 ( l)$r H,c o F o

o0+r o {J oFl O .'.1 Cl O t{.1r{ 3€ E-C dt !X C E-

td (d q) .cHA 'c O

.OO.u+J..F. tlH O bou| E_c

E bi i , l E g 3(!P (Hl{COU|F (n C tN..r ' - rb0 . r{ t4 t rOo o P o C-oi{ ..1 tr .'{ .'-f O Exs 3 U)-c x'qF O >,x ' oa t .o_{ oo ?h d.c xO. ' t t t r {J-Oh-r E <D cd

s .$f lE,gB(/).db0oo5+r

Fl.-l (/) {J l i (6strodPo.+) t r H C O.- t

c. ' l O O A.Fi tr tN,C orJ O tr. , iS, h h o ' r ' r. . l i l ] 'qooo(nfl \- +J .Ft O P .,-{

208

<u<UQUO-

!U<EOJ

-AOUQG-*-QaF-

OOA-u@4a<JxoJ<o

-u-<9o9=oadF-a-<E-AOeqJO=O<J€<-eJ=-d=oz<JJ-<=UOG<<o-)==a

' .>.a.a. l

dtHsirottnlu

209

PHYLOGEM OF THE OLIGOCFIAETA

namely Hotnogaster, MLcroeLtaetus, Pontoseoler, GLossoseoLen, Hemigastro-dyilus and Kynotus. This grouping represents a restriction of the Lun-bricoidea s. Jamieson, L97Ld, to exclude the almids, CyiodyiLus, Spay-garnphiLus and BiaadrLlus.

The fact that the strperfanily Megascolecoidea (excluding the ocne-rodriles) is the aponorph sister-group of the Lunbricoidea, with the Lat-ter intervening between then and the almid complex is of the greatestinterest. I t suggests the forrner existence of a conmon ancestor inter-nediate in form between the negascolecids and the lunbricoids. Thisancestor, i t is indicated, had no Lateral Lines, possessed a nult i layeredcli tel lun, with snall yoLked eggs, was holandric with 2 pairs of nalepores united in segnent L3, (unspecial ized) ovaries ( in XIIIonly?) long ser ies of prostate- l . ike setal g lands, last hearts in XI,intestinal origin in XIV and adivert iculate pretesticular spernathecae.A special relationship between the Megascolecidae and non-alnid I 'glossos-colecids'r was anticipated by Janieson (1974a) in a discussion ofHieknanielLa which is alnost unique in the Megascolecidae in having anintestinal gizzatd. A11 the shared characters noted between negascolecidsand glossoscol.ecids are now seen to be aponorphies in Hennigts terminology.Inclusion of the GlossoscoLecidae and Lunbricidae in the superfanilyLumbricoidea and of the Megascolecidae and EudriLidae in the superfanilyMegascolecoidea was upheld and recognition of the suborder Lunbricinaas a natural grouping was endorsed. The Al.luroididae, included in theLumbricina by Brinkhurst and Janieson (1971), reLated the suborder toother haplotaxid suborders, especial.ly the HapLotaxina but they areexcLuded fron the Lunbricina in the present study.

Whereas the coverage of the Alnidae in this investigation incLudesall known genera of that fanily, only a few genera of the Megascolecidae,representative of the Acanthodrilinae, the three tribes of the Megasco-lecinae sensu nihi, t97La, and the Ocnerodri l . inae, are included. Thephylograrn and phenogran for the Megascolecidae cannot therefore be regardedas more than heurist ic. Interesting aspects of this port ion of the phy-lograrn are the dist inctness of the OcnerodriLinae, here recognized, asdid Gates long ago, as a separate fanily, confirnation of the Eudrilidaeas the plesiomorph sister-group of the Megascolecidae, and the very dis-t inct, highly aponorphic status of the three representatives of the tr ibeMegascolecini Janieson, L97La, Pheretima (type-species fron Tahiti butprobably Indonesian or Malaysian) , SpeneerLelLa (Australia) and Lanryito(India), the close grouping of which lends sone support to recognit ionof this tr ibe. Another heurist ic grouping is that of two species ofDLehogaster (Africa and Asia; Megascolecinae, tribe Dichogastrini Jarnieson,L97la) with the neronephrlc DLplotrema sp, (Australia; Acanthodrilinae,tribe Acanthodrilini) and Octochaetus (New Zealand; tribe 0ctochaetini)as the plesiomorph sister-group of the Megascolecini. This groupingtends to confirm the authorfs suspicion that those Dichogastrini withacanthodrilin nale pores (India and Africa) are descended fron Acantho-drilinae and are distinct fron dichogastrins with negascolecin nalepores (Oriental and Austral ia). Such negascolecin dichogastrins arepresunably descended fron negascolecin holonephric forms (Megascolecinae,tribe Perionychini Janieson, L97La). The latter view is supported bythe sister-group relationship between the megascolecin dichogastrinDLgaster, on the one hand, and the perionychins Hetev.oporodrLlus

zLO

50 Pxenrrrnr48 Lmprroq7 SpencsnreLLaqq DtcxoorsrEn o.45 Drcxoelsren s.58 0crocxaerus49 CEueRreuun57 DrplorRenn sp,45 Drarsren s.42 Cnyproonrlus n.40 HEtEnoponooRrr-us36 Dr plornEr,u r.35 ltltcnoscolex4l Dtponocnreu46 lleerscolroEs59 PenlouYx32 Eurennra31 0cneRoon r lus3q Nersntounrlus33 Euon r lus30 Etserreur29 Lumsnrcus2\ Dtponoonr lus23 HEillonsrnoontrus22 Vlenvse25 Aruoscolex2l HoRr'roe nsren17 lhcnocneetus26 KomrRerrora9 SpaneaHopn t uus

16 Cl-r-ronrlus u,15 Clt . r ronr lus s,14 GlvpnronrrocnrusB Dntrocntus H,12 DRruocRtus l .10 Ar-nr11 Glvpx r on r lus28 Cnroonrrus27 Luronnrrus5 SyHeeHonn I uus4 Suloenrn3 BnturxuRsrr l2 Aulunoroes1 Haprorrxt s8 Brwaonrr-us

t9 Porroscolex18 Grossoscolex20 Kvrorus7 DesNoelsreR6 lt loHruteasren

Fig. 3. Phenogram for al l famiLies of opisthoporous ol igochaetes(HaplotarLdd. TTre phenogram was constructed by nearest neighbourclustering of the sini lari t ies indicated in a matrix of coeff icientsof sinilarity. These coefficients were obtained by sunrnating positiveand negative natches for the characters employed in preparation of thehennigram shot{n in Fig. 1. Positive matches were synaponorphies andnegative natches were synplesiomorphies in Herurigian tenns but inphenetic terfig these were nerely natching character-states requiringno considerat5.on of evolutionary change. The phenogran differs fron aneo-Adansonian dendrogran insofar as selection of characters was influencedby consideration of their ut i l i ty for Hennigian analysis. For treatmentof nult istate characters see text.

zTL

PHYLOGENY OF THE OLIGOCHAETA

and ClzJptodyLltrc on the other, all three from Australia. The diphyleticnature of the Dichogastrini, and origin of African mernbers frorn acantho-driles, is further indicated by the absence of perionychins fron Africa.

While the diphyletic origin of the Dichogastrini energes clearly,the interrelationship of acanthodriles and perionychins suggested areperplexing. The classical view of origin of the perionychin arrangementof male pores (a pair of nale and prostatic pores on XVIII) fron theacanthodri l in condit ion (nale pores on )UIII, prostate pores two pairs'on XVII and XIX), a view questioned by Janieson, (1963), is not strpported.This is not to deny the probabil i ty of origin of the nicroscolecin, balan-tin, or sone megascolecin conditions fron the acanthodrilin condition.The phylogran is not inconsistent with the revolutionary view that theacanthodril.in condition nay have originated by replication of prostatesfron the perionychin condition. In view of the occurrence of sexpros-tatic forns (e.g. DLehogaster. darnonis and PiekfordLa) in the Dichogas-trini and Acanthodrilinae and of forns with six or more prostates inthe Ocnerodri l idae, a nult iprostatic origin of the Megascolecidae, givingthe acanthodriLin and perionychin conditions independently seens morelikely, however. The alternative, origin of perionychins fron acantho-driLes with honeotic displacenent of the anterior prostates to segnent)ffIII (a displacement conmon in New Caledonia which harbours the rnostprirnitive known perionychin genus) wiLl be discussed in a later work.

Separation of DLplot?erna fyagLlis, the type of this AustraLian genus,frorn the neronephrLc DLpLotrema sp. when rnorphologicaL studies indicatethat the two aTe very closely rel.ated is not easiLy expLained; nor isthe interpolation of the Indian CeLerLeLLa (tribe Megascolecini) betweenit and the acanthodrilin group. NevertheLess, the DLplotrema sp, renainsthe phylogenetic nearest neighbour of DLpLotrema frugilis,, and-SpeneerLeLLais the phenetic and phylogenetic nearest neighbour of Celeriella. Theposition of Mtey,oseoLes (Notiodyi,Lus) geongtaruts, (S. Georgia) the type-species of the Acanthodrilinae, separated fron other acanthodriles bythe three Austral ian Megascolecinae is aLso diff icuLt to interyret. I tis noteworthy that the phyletic and phenetic nearest neighbour ofM, (N.) georgiarnn is DLplotrema fragiliso DLpLotrema being placed inMLcroseoLer as a subgenus by Janieson (1974c).

The relatively isolated position of MegascoLides austm,Lis is con-sistent with the authorrs view that this species has acquired longitudinalureters (placing it in the Megascolecini) independently of the trueMegascolecini as is apparently the case in sirnilarl.y characterized (butnot conputed) species of Crgptodnilu,s (see Jarnieson, 1974a),

It wil l be noted that the phyLograrn calls into question divisionof the Megascolecidae into the subfanilies Acanthodrilinae and Megas-colecinae. The subfanilies are retained for the present pending a thoroughanalysis of the fanily. I t nay be possible taxononically to retain themby including those dichogastrins with acanthodri l in male pores in theAcanthodvilirwe.

The lack of cohesion of the Perionychini (cornputed for PerLonyr(India) , DLporoehaeta (New Zealand), HeteroporodrLLus and Cz'gptodriLus(Australia) nay reflect the fact that this tribe is based on the syn-plesionorphy holonephridia, though in combination with the relativelylponorphic but not unique character nale and prostatic pores on XVIII.

2L2

B.G.M. JAMIESON

However, the special. relationship between Cnyptodrilus and Heteroporo-driLus suggested by Jaurieson (t972a) is borne out by both nearestneighbour relationships. Those Dichogastrini with male pores on XVIIIwould renain a valid group within the Megascol.ecinae though removal fromthis tribe of the, mainly African, acanthodrilin members (including thetype-species of DLelrcgaste ) would require the negascolecin portion tobe renaned. The tribe Megascol.ecini would appear val"id for forms withnal.e and prostatic pores in XVIII lacking the dichogastrin or periony-chin conditions of the nephridia, though sone nembers which have conver-gently acquired a megascolecin e$cretory system nay have t-o -!e excluded.Meronephric, non-dichogastrin nenbers of the genus CryptodrtZas possess-ing ureters have already been excluded (Janieson, L974a).

LwnbrLeida,e

Within the Lumbricoidea the Diporodrilidae are cladistically, inphylogenetic and phenetic nearest neighbour relationships and in thephenograrn, closest to the Lunbricidae, of which tJrey form the plesio-lnorph sister-group. Significant diporodrilid synaponor?hies with theLunbricidae are: presence of tubercula pubertatis; Location of nalepores on XV, wel.l anterior to the clitellun; Presence of an intestinal.gizzard preceded by a crop; the intestinal t1ryhlosol.e; possibly single-itringed ovaries; adiverticul.ate speflnathecae in the testicular seg-nents and the presence of adiverticulate nephridial bl.adders. Thesymplesionorphy absence of an oesophageal gizzard also seems worthy ofrbntion. IL is here considered that the singJ.e genus , DLponodrLLtn,nust only be assigned to a subfamily, the DiporodriLinae, within theLtrnbricidae. Presence of a longitudinaL ureter interconnecting thepostcl i tel lar nephridia is no obstacle to inclusion in the Lunbricidaeis a sirnilar duct occurs in AlLolobopltow. anbi,pae. The unique auto-apornorphy of the Diporodrilide is the paired coel.onic pores contrasting*ittr ttre nedian urpaired pores of the Lr:rnbricidae. They may have evolvedfrorn nedian dorsal pores or may represent an independent developnentfrom an ancestral imperforate condition. fire prior groqps in the phy-logeny (Haplotaxidae through Al.nidae) lack dorsal pores. Bouche (1970)considered that the Diporodrilidae could be derived fron either theLunbricidae or the Homogastridae but later (1972) decided that originwas from the Hormogastridae.

AiLo s eole ei' dae and Horrno gas trtdn e

Linkage of Ailoscolen with Horrrcgaeter is heuristic. AiLoscoleragrees with the Hormogastridae s. BouchS (L972; including Horrnogastez',HemigastrodrLlus and Vlgnysa) in significant respects: setal intervals,presince of longitudinally grooved genital setae (personal observation),presence of nephridial caeca, probably in having a single egg-str ing tothe o.raty (uncertain in AiLoseoLes in which the ovary is certainly notfan-shaped, personal observation), location of last hearts in XI, inpossessing two oesophageaL gizzards (as in Vi'gnysa; HorTnogaster andttemigastzodyLlus have three) and in location of the first gizzatd inVI; 6ut the nale pores are far posterior in AiLoscoLen (7/3 XXII) com-pared with Hoymogaster, HemigastzodYilus and Vigrtysa ()U). Because

2L3


Horrogastaz' links with, and is the nearest neighbour of, AiLoseoLeuboth phyletical ly and phenetical ly, i t is here proposed to include itin a fanily Hormogastridae with the three other genera nentioned, regard-ing it as aponorph in being the onLy hormogastrid in which nale poreshave nigrated posteriorly of segment XV but plesiornorph in having spel-mathecae in, but not behind, the testis-segnents. Subfanil ies recognizedare the Hormogastridae (Horrnogasten" HenrigastrodrLlLt's ), Vignysinae(Vignysd, both incLuded by Bouche , L970, and Ailoscol.ecinae (AiLoseoLeu)proposed as a separate fanily by Bouche, 1969.

Microchaetidae

MLctoehnetus appears to be the plesiomorph sister-group of the fam-ilies Lunbricidae s. lat. and Hormogastridae (represented by HorTnogasterand Ailoeeoles). Sonewhat unexpectedly its phyletic and phenetic nearestneighbour is Ailoseoles. Ttre view that it is the sister-group of theGlossoscolecini s. Jamieson, L97Ld, is not confirmed. It is here pfop-osed to recognize the fanily Microchaetidae for MLerccVtaetus and (notcornputed) IrLtogenia, differing fundanentalLy frorn the Microchaetidaes . Gates in exclusion of the aLnids.

Kynoti,dae

The MalaysLan Kynotus was shown as the sister-group of the honno-gastrins, microchaetins and gl.ossoscolecins in the fr intuit iverrphylogeny

of Janieson (1971d). I t maintains this posit ion in the present phylo-genetic study.

Hennig's view that affinities should be determined only on the basisof shared aponorphies is nowhere more cl.earLy supported than in the phene-tic relationships of Kgrnkts for, when negative (pLesiomorph) natchesare included, this genus is far displaced from hormogastrids and nicro-chaetids in the phenogram and has as its nearest neighbour (Table 5)the eudriLid NemertodniLrc" affinities which seen inacceptable . Kynofusis here assigned to the nonotypic faniLy Kynotidae and, as formerly,lies within the Lr:mbricoidea.

GLossoseoLecida,e

GLossoscoler and Pontoscoler, representing the S. Anerican glossos-colecids, here given farni l ial rank as the Glossoscolecidae s. str ict.in accordance with the views of Gates (L972), are phenetic and phyleticnearest neighbours. The Glossoscolecidae s. Lat. is a large and hetero-geneous group, with, for instance, spernathecal Locations varying fronpre- to post-testicular, and may be found on further analysis to requirelubdivision into subfamil ies or even further fani l ies.

Komarekionidae

The plesionorph sister-group of the superfanilies Megascolecoideaand Lumbricoidea in the phylograrn is the nearctic Konarekionidae of Gates,!974, but Komarekiona has MLenochaetus as its phylogenetic and Sparga-nophilus as its phenetic nearest neighbour. It is here included in the

2L4

B.G.M. JAMIESON

Lunbricoidea as a distinct fanily. Bouch6 (personal conununication)considers Komarekiona to be an rrAiloscolecidfi and there are, indeed,str iking sini lari t ies including location of the rnale pores on XXII, butKomareki,orw has only a single oesophageal gLzzatd, in VI, (2 in VI-IXin Ailoscoleu) and intestinal origin in XIII, not XV; lacks nephridialcaeca and intestinal typhlosole and has pretesticular, not testicular,spernathecae. The spermathecal difference seems significant though theother differences are known to occur within good genera in, for instance,the Megascolecidae, AiLoscoLen Lacks the two pairs of lateroparietaLvessels of Komarekiorta (personal observation for 4. Lacteospwnosus)and, coupled with the spernathecal difference, this suggests that theKonarekionidae should be retained.

ALmidae and CriodrLli'dae

The phylogram gives strong srryport for recognition of a groupingthe Alninae (Janieson, 1971d) or Al.nidae (Ljungstron, t974; Jamieson,L976). At least fani l ial rank is justi f iable for this assenblage.Indeed, if Herurigfs view that sister-grol4)s nust be given co-ordinaterank be accepted, the alnid group must be regarded as at least coor-dinate with and perhaps of higher rank than the combined negascolecoidand hmrbricoid superfanilies. Thus the alnids and their plesionorphsister groqp CriodyiLw nay be regarded as comprising the familiesALmidae and Criodrilidae of a singl.e superfamiLy, the Almoidea. Thesole genera of the ALnidae, al l represented in this study, areCaLLidriLue, DrLlocritn, Glyphidriloenius, ALna and GLyphidrLlusas noted by Janieson (1971d, L976). Origin of the Lunbricidae fron theCriodri l idae, (vide Janieson, 1971d) is not supported.

LutodriLidne

The genus Lufudyi.Lus, rmique in the earthworns in having ten pairsof testes, in XII-XXI, and a Lurnbricoid s. Gates, is a sister-groupof the almoid h:mbricoid-megascolecoid clade (ignoring the questionablyplaced Vignysa). According to Hennig's principles it therefore warrantsi rank higher than superf,arnily. Its phylogenetic nearest neigfubour is,however, GLyphidrilus and its phenetic nearest neighbour is CaLlidtilw'both almoids. While the author accepts the evidence of McMahan (1976)that presence of testes in XII-XXI is normal for the single known species,the affinities of the genus have been computed as though it had the nor-nal megadrile arrangenent of testes in X and XI. LutodrtLrc has clearlyinterpolated ten segnents, the last eight of then testicular, anteriorto this normal location but to have recognized this for computationalpurposes would have nade other organ systens consequentially non-congruentwittr ottrer megadriles. If the ten additional segments are deducted,conparison with GLyphidrilus and CaLLi,drtlas is facilitated. Aponorphieswhich Lutodyi,Lus shares with these genera which are of particular noteare: (a) the quadrangular posterior cross section; (b) dorsal anus;(c) copulatory alae (with GLyphi.drilus onLy); (d) presence of a subneuralvessel, (e) an intestinal typhlosole, (f) nult iple posttesticular sper-mathecae and (g) last hearts in XI. It is here proposed, a posterLorL"to weigh the alnoid apornorphies (a-c,f) which are rare or unique within

2L5


the oligochaeta, to the extent of including LutodrtLus as a third fam-ily in the Almoidea. This family stands apart fron other alnoids inhaving single-str inged ovaries.

0cnerodrilidae

A noteworthy feature of the renainder of the phylogram is exclusionof the two ocnerodriles, 1enerodriLw and Eukeruia, fron the Megascole-cidae by intervention of the Alnoidea and Ltrnbricoidea. The ocnerodrilesare, however, linked with the Megascolecidae in the phenogran (Fig. 3)in which the two are mutual nearest neighbours whereas the phylogeneticnearest neighbour of Eukerria is the rnegascolecid MLeroseoler geongianlLs,If credence is to be attached to this portion of the phylogran it wouldhave to be accepted that the 0cnerodrilinae arose fron early megadrileswhich, like present day ocnerodriles, and Spanganophi.Lrc' *hi9! they arecladistically near, had the Last hearts in XI and (Sparganophilus) a veryshort oesophagus and netamericalLy repeated prostate-like glands. Theseglands persist, supernumerary to true prostates, in several ocnerodri legenera.-

In the phyl.ogran the ocnerodriles are the plesionorph sister groupof the megaslolecoid-Lr.rnbricoid-anoid conplex. A close cladist ic aff in-ity of ocnerodriles with the alrnoid-lunbricoid conpl.ex has rarel-y beenpr-oposed but it has been noted (Jamieson, 1965) that the ocnerodrilesWeiatogenta Laeuu,rn and IV. pdnenaerlais resenble glossoscolecids in havingthe nale pores anterior to the niddle of a long cl i tel l .um, unlike negas-colecoids.

If the ocnerodriLes were accepted as cladist ical ly very dist inctfron the Megascolecidae and fron the sister-group of the Latter, theEudrilidae, the strpposed origin of eudriLids frorn ocnerodriles couldnot be upheld, ocnerodri les can no longer be categorical ly associatedwith the Megascol.ecidae and Eudrilidae and separate fanilial status, asthe family Qcnerodri l idae, advocated by Gates (e.g. 1959) and by Sins(L966, tg-67), though these authors did not suggest alnoid-lunbricoidaff init ies, appears justi f ied. I f ocnerodri les were accepted as beingcladist ical ly dist inct fron the Megascolecidae, the development of tub-ular prostates and in sone ocnerodriles (e.g. EukerrLa) of the acantho-dri l in condit ion of the nale terminalia (prostate pores in XVII and XIX,nale pores in XVIII) would have to be regarded as rernarkable anticip-ations or convergent development of prostates in negascolecids and ofthe acanthodrilin arrangement of these in acanthodrilin and Afro-Orientaldichogastrin Megascolecidae. It wouLd also necessitate acceptance thatprostate-l ike glands ( ' tsetal glandstr) capable of giving rise to prostatesiecreting into the spernathecae of the concopulant were persistent throughsuccessive ancestral forns which gave rise respectively to the Sparga-nophil idae (and earl ier groups such as the Biwadri l idae), Ocnerodri l idae,Allnoidea, Lurnbricoidea and Megascolecoidea. Prostates or prostate-l ikeglands are seen in some or most menbers of all" of these groups.

In the Light of the present phylogran we may envisage ancestralforrns of the Megascolecidae and Eudrilidae with long series of setalglands (precursors of true prostates which latter are here terned meta-

frostates) and with a facultative tendency to nodify the ectal ends ofih" rperm ducts (precursors of euprostates). From such ancestors it

2L6

B.G.M. JAMIESON

could be surmised that negascolecids originated by ernphasis on deveLop-nent of metaprostates fron the setal. glands, though not always to theexclusion of nodification of the vasa, whereas eudrilids developed byadopting the euprostate, losing the setal glands, and foregoing thedevelopnent of the prostate. These conclusions still leave the questionwhether ocnerodriles are as distinct fron negascolecids as the phylogranand phenogran indicate, at least for the smaLl sample taken. If theyare distince, convergent development of metaprostates and of the nicros-colecin and acanthodriLin conditions in ocnerodriles and negascolecidswould have to be accePted.. Before such convelgence, or parallelisn is dismissed as incon-ceivable it nust be recognized that sorne such convergences have to beaccepted whether ocnerodriles are considered to derive from (1) an in-mediate ancestor of the Megascolecidae-Eudrilidae or (2) a corunon ances-tor of the megascol.ecoid-lurnbricoid-al.moid assenblage. Thus if ancestor(1) were a proto-ocnerodrile (with or without the auto-aPonorphy of extantocnerodri le!, presence of calciferous-t issue or glands in IX) i t wouldhave to be accepted that it gave rise to forms in the MegascoLecidae-Eudrilidae with a wide variety of arrangements of the nale terminalia(acanthodrilin, as in the ocnerodrile fi,tketwta; nicroscolecin, the com-mon ocnerodrile and eudrilid condition; negascolecin and bal'antin, tonane the chief variants) and that either the acanthodrilin or the nicros-colecin conditions of ocnerodriles of which the former is infrequent andthe latter usual, developed as a convergence in relation to the megas-colecids. Parall.el developnent of these arrangenents, which appear fromtheir frequency to have been highly successful, is not nore surPrisingthan the evolution of the wide diversity of arrangements (which seern tohave had Linited success) exernplified by @rdi.odri.Lus" Narm'odrilus,Thatonia" Deecania and other ocnerodriles which have neither the acan-thodrilin nor the negascolecin arrangenent.

If the ocnerodriles were divorced further fron the Megascolecidae-Eudrilidae (condition 2) as in the phylogrann, and the proto-ocnerodrilewere at the root of the negascolecoid-lurnbricoid-alnoid assenblage,it would have to be postuLated either that the proto-ocnerodrile whichgave rise to the Ocnerodrilidae on the one hand and the assemblage on.ihe other had setal glands unmodified as prostates or that it had aLreadymodified at least sone of these as netaprostates. In the forner casethe prostates and shared arrangements of ocnerodriles and negascolecoidswould be examples of convergence. In the latter case it would be in-ferred that prostates had been lost in aLnids and lumbricoids lackingthen while setal glands were not infrequently retained. Origin of pros-tates and their arrangenents in ocnerodril.es independently fron that innegascolecoids cannot be categorical ly rejected. Alluroidids, especial lyAr.l,nkhtastia, (vide Jarnieson, 1968b) denonstrate origin of sirnple tubularprostates irr a lineage which now appears to be remote fron the Ocnerodri-t ia.u despite sini lari t ies cited by Jarnieson (1968b) and Ljungstron(1974). in Brinkhtastia each prostate opens with the corresponding vasdeferens, into an atrial chanber much as in the ocnerodril'e genusNannodrilus. Alluroidids could thus be taken as evidence either forindependent origin of netaprostates or for the existence of prostatesin eirly ancestral worms which gave rise to nodern alluroidids, tononil igastr ids and to the Lurnbricina, including ocnerodri les.

2L7


Returning to the possibiLity of direct ocnerodri le-negascolecoidaffinities, a conmon ancestor of the Megascolecidae and Eudrilidae maybe assigned plesionorph features of either or both fanil ies, o.g. shortoesophagus with intestine conunencing in XII, absence of spermathecaldivert icula, and location of last hearts in XI. Only calciferous mod-ification of the oesophagus in IX (or IX and X) would be necessary toproduce an ocnerodriLe. Although these plesionorphies would also beattr ibutable to ancestors of the megascolecoid-lunbricoid-alnoidancestors, the latter ancestors would have had nale pores on XV (oreven XIII) as in Lr:mbricids, hornogastrids and Cvi'odrtLw, It is onthis point, position of the nale pores, that convergence rather thannonophyly, of ocnerodriles and megascolecids with regard to nale terni-nalia is difficult to accept. Although it nust be accepted that sonenegascolecid-ocnerodrilid arrangenents of the terninalia have evolvednore than once, the location of the pores of the vasa deferentia on)UII and/or XVIII (though also seen in sone lunbricoids and alnoids)taken in conjunction with unique sirniLarities of the prostate arrange-nents (aLbeit by paral lel isn) in the two faniLies incl ines the authorto consider ocnerodriles as nonophyletic with the Megascolecidae-Eudri-l idae and as l" i t t le changed representatives of the ancestors of the lat-ter coupLet.

SpangarnphiLidae

Janieson (1971d) pl.aced SpargarnphiLw as a nonotlryic subfanilyin the Glossoscol.ecidae s. lat. on the grormds that glossoscol.ecidaffinities were indicated by the intraparietal sperm ducts (a conditionshown by Pickford, L976, not to be r:niversaL in ttre glossoscolecids,however) and a group of characters which individualLy were not restrictedto or even general in the famiLy, Oiz. the extensive cLitel lun, and thehistology of this, the intracl. i tel lar nale pores with longitudinal tu-bercuLa pubertatis; the transversely cicatriced setae; the tendency totransverse multiplication of the adiverticulate spermathecae; the pos-terior dorsal groove and the dorsal anus, seen in the Almidae which werethen included in the Glossoscolecidae. These character-states' with thepossible exception of the intracl i tel lar location of the nale pores,lnd presence of p"ostate-like glands, are apomorphic, a fact furthersupport ing recognit ion of glossoscolecid ( i .e" lunbricoid-alnoid)afl init ies of SpargarnphiLus. Possession of prostate-l ike gLands, sharedwith Kynotus but also with BiuadyiLrc (then pLaced in the Glossoscole-cidae) was a further sini lari tY.

It has been pointed out (Jarnieson, 1971d) that in Spargaraphilusthe tendency to backward movenent of the naLe pores is well developedand that the brain has noved back to segnent III but that evidence ofthe prirnit ive origins of the genus relative to many rfglossoscolecids' l

were seen in the exceptionally short, lxlspecial ized oesophagus, in ab-sence of a subneural vessel, and retention of precardiac connissuralvessels, SpargarnphiLus was shown to resenbLe the Ltrnbricidae in havinga single-sti inged ovary, intracl i tel lar tubercula pubertatis, a saddle-shaped clitellun and no oesophageaL gizzatd. It was considered thatSpangarnphilus was a litt1e nodified representative of the ancestralGtosloscolecidae s. 1at . and, accept ing a glossoscolecid or ig in of the

2LB

B.G.M. JAMIESON

Lgnbricidae, it was thought probable that the lunbricid-like features ofSparganophi,Lrc were an example of parallelisn in related forns rather thanconvergence of unrelated forrns.

gnodeo (1956) saw S"ptangarnphiLtts as representative of the ancestralstock of the Microchaetidae; Hormogaster (which he placed in a subfan-ily of the Microchaetidae); the Lunrbricidae; and Criodniltn, ALma andDz'iLoeri,w, which he considered night be placed in the Lunbricidae orin a separate farni ly. The GLossoscolecidae s. str ict. were shown asthe sister-group of SpargarnphiLus and its above-mentioned supposeddescendants. Kynokte was assigned to the subfamily Microchaetinae.Renoval of ALma and DrLlocrius fron the Microchaetidae is endorsed in thepresent account though the other affinities recognized differ signific-intly frorn those deduced by 0nodeo. Yanaguchi (1953) ascribed a relat-ively advanced position to Spargartophilw.

In the phyl.ogram origin of SpanganophiLw fron the basal stock ofthe Glossosc-oLecidae s. Lat. (Lunbricoidea, Alnoidea and BiuadrLLus) isconfirmed if the intervention of the ocnerodrilidae, questioned above,be ignored. It now seens acceptable that Spangarnphiltrc is descendedfron the root of the alnoid-Lunbricoid (rnegascolecoid) complex as st4rportedby its nearest neighbour relationships which are phylogenetically withiomarekiona and Lutod.riLw, (fringe members respectiveLy of the Lumbricoideaand Alrnoidea) and pheneticaLly with Komarekiona,.

It is particularl.y noteworthy that Longitudinal tubercula pubertatisseen in SpanganophiLus but very rare elsewhere in the Ol.igochaeta, occurin all th; pnbricoids computed with the exception of Kynotuts (EisenielLa'

LwnbrLcu,so DLpotodz,iLrc, AiLoseoLex, Hor'rnogAetet" MLercehaetus, Pontos-coLer, GLossobeoLen, HewLgastrodrLLus and Komarekion'd and that tuber-cul.um-like structures occur in the almid CaLLi.drLLrc. The longitudinalalae of the al.nid GLyphidrLlus are probabLy honologues of tubercula puber-tat is.

In sunnary, SpatgaraphiLw is here ernvisaged as being descended froma precursor of the alnoid-Lurnbricoid assemblage, with intracl i tel lartuLercula pubertatis, maLe pores on the rnult i layered cl i tel lun, in XIII,prostate-like setal glands in rnany segnents thoygh lacking prostates,iast hearts in XI, intestinal origin in IX, lacking a typhLosole andprobabl"y agiceriate; holonephric but perhaps lacking preclitelLar nephrida(or is ttrii aUsence nerely a derived character of haplotaxids, alnoids and

BpoogonophiLus in response to an anphibious existence, as in the negas-

"bf"lia PontodyLlus?); holandric; and metagynous. Whether the ovaries

were single-stringed, or webbed, or as in Cri.odrLLrc Lacking linear strings,is less Lertain. A11 the lumbricoids computed with the apparent exceptionsof MLey,oeVnetu,s, PontnseoLer" GlossoseoLex and Kynott't's, have single-str inged ovaries. This nay be a device for faci l i tat ing placenent of

the eggs in the ovisacs (Gates, L976) and parallel development of this

featuiE has probably been a strong trend throughout the Lunbricoideas. nihi presunabll l start ing from an unspecial ized criodri l id condit ion.Gates considerea ttre single-str inged ovaly diagnostic of his restr ictedLunbricoidea.

In Hennigian terms Spaz,garnphiLrc must be accorded a rank higher than

superfanily but it is here placed in the new superfanily SparganophiLoidea.

2L9


BiuadriLida.e

The nonotypic Japanese fanily BiwadriLidae (Biwadrilinae s. Jamieson,1971d) appears in the phylogran to form a nonophyl.etic grouping with thesparganophiloid through negascolecoid conplex. Its sole species, BiuadrtLusbatTtgbates, was removed fron the genus CrLodyLlu,s by Jamieson (1971d) andassigned to a subfanily Biwadri l inae of the Glossoscolecidae s. lat.Features warranting placing it in the Glossoscolecidae but now consideredto be of suprafamil ial signif icance are: the extensive, nuLti layeredcLitel lun; the location of nale pores near the posterior l init of theLatter; the presence of setal glands resenbling those of MLcnoehaetus;and the intraparietal sperm ducts. Affinity with SparganophiLus wasconsidered to be indicated by presence of two pairs of lateroparietalvessels, their origin in segment XIV, and the absence of a subneural ves-seI. This relationship is rmderlined by the fact that SparganophiLus isits phenetic nearest neighbour though its phylogenetic nearest neighbouris MLerochaetus.

Reported presence of lateraL l ines, absence of spernathecae, theanterior Location of the brain ( in segnents I and II) and other peculiar-ities necessitated placing it in a separate and apparentl.y very prinitivesubfanily of the Glossoscolecidae. As it now appears as the plesionorphs i s ter- grotrp of the sparganophi loi d-almoid- lunbri coi d-negas co le coidcomplex, i t deserves, i f Hennigian principLes are str ictLy fol lowed,a status above that of superfanil.y. It is proposed here to give it thestatus of a monotypic family within its own superfanil"y in the suborderLunbricina.

Moniligastridae

0f the three orders of the Oligochaeta recognized by Brinkhurst andJanieson (1971d), the MoniLigastrida and Haplotaxida are represented inthe present study and it is seen (Figs. t q 2) that separate ordinaLstatus for the Monil igastr idae (0riental) is not justi f ied. This reduc-tion in rank has already been advocated by Jamieson (L977a) on denons-tration that nonil igastr id testis-sacs are not intraseptaL prol i ferations.On the other hand the phenetic analysis (Fig. 3) depicts the nonil igastr idsas a highly distinct group as shown in a previous phenetic study (Janieson,1968b). Their dist inctness is al,so indicated by the fact that they arephylogenetically and phenetically nearest neighbours and do not constitutethe nearest neighbours of any other OTUs in the study. The reinterpret-ation of the nature of the nonil igastr id testis-sac reveals that each ofthe one or two pairs of nale pores opens at the posterior border of thesegnent behind the corresponding testes. This init ial opisthopore con-dition and the extremely plesionorph retention, in Desmogasfur, of twopairs of nale pores in separate segments are limited elsewhere in theHaplotaxida to the Haplotaxidae. Moniligastrids are therefore extra-ordinari ly prinit ive in these respects, as in the possession of a uni-layered c1ite1lun (queried by Gates, L972, but here confirmed forMoniLigasten troyi) and the correlated large-yolked eggs. However, asGates (L972) has stressed, they have an unusually large conplenent ofunique characters which in Hennigian terns are to be regarded as apomorph.

220

B.G.M. JAMIESON

These include suspension of thd testis-sacs on septa, the presence ofenterosegnental. organs, the ovarian sacs (see also sorne eudrilids), thenultiple postoesophageal gizzatds, the absence of peristonial attachmentof the prostoniun and (not computed because of Lack of knowledge in othergroups) union of hearts before conmunication with the dorsal. blood vessel.These auto-apomorphies, while unifying the MoniLigastridae, bI virtue oftheir uniqueness do not detemine the position of the family relative toother OTUs in the phylogran. It is the apomorphies shared with variousother groups which do this.

The MoniLigastridae are seen as the plesionorph sister-group of thecomplex BiuadrLLus through Megascolecidae (the Lumbricina). Ttre Monili-gastridae will therefore constitute the suborder Moniligastrina. Thelniee evolutionary step at this dichotony is fusion of the two vasadeferentia of a side in the Lunbricina. In Desmogaster the two vasarenain sepalate, giving two pairs of mal.e pores, whiLe Ln MoniLigasterthere is onLy one pair of nale pores, not because of fusion but becauseloss of the anterior pair of testes has occurred.

SyngenodriLidae

SyngerndrtLus, inrperfectly known from a single Ethiopian specirnen,was pliced in the Moniligastridae by Stephenson (1930) as a nonotlpicsubfamily but was Later transferred to the Al.l.uroididae (Pickford, 1945;Gates, 19451 Janieson, 1968b, 1971d). Association with al l .uroidids mayhave been unduLy influenced by the sympl.esionorphy singl.e-Layered clitel-lun and the relatively primit ive Location of the naLe pores' on XIII.Its nearest neighbour phyleticalLy is Pontoseoles (Table 3) and phene-tical ly Ls ALluyoidBs (Tabte 3 and Fig. 3). Notabl.e aponorphies ofSyngenodrilus distinguishing it frorn hapl.otaxids and alluroidids aretire-presence of intracl i teLl.ar tubercula pubertatis, of three,pairs ofprostate glands (in XI, XII and XIII) of gizzards, in VIII and IX, ofnephridial bladders and apparentl.y the loss of lateral lines. It is pos-silte that fusion of the nale pores indicates that it is a survivor ofthe earliest stock of the Lunbricina, al.one retaining the single-layeredcli telLun and Large yolked eggs. It is here proposed to retain i t as anonotypic fanily in the ALluroidoidea notwithstanding the paraphyl.eticnature of the superfanily so constituted in strict Hennigian terns.

ALLtuoi&Lrne

The alluroidids StandeyLa and Bz"tnkhuvsti.a form the plesionorphsister-group of SyngerndriLw through the Moniligastrina and Lunbricina.It is unlikely that the Desmogastez' condition of two pairs of nale poresis a reversion and it is therefore inferred that the ancestor which gaverise to the Alluroididae and Moniligastridae sinilarl"y had 2 pairs ofpores with fusion occurring in Syngenofu,iLus independently of the Lun-Lricina. The interesting conclusion therefore emerges that theAlluroididae, characterized by a single pair of maLe pores, on XIII,and a single pair of testes, in X, (the proandric condition) have neverfused the vasi deferentia and it nay reasonably be supposed that theancestor of the Alluroidoidea had two pairs of rnal.e pores, in XI and XII,as in Desmogaster and sone Haplotaxidae, and that loss of the posteriortestes proUibty occurred before movement of the renaining male pores from

22L


segment XI to XII I .The affinities of the three alluroidid genera here computed have

been discussed by Janieson (1968b, 1971d). The phylograrn supports theview there stated that the S. African Standerta and S. AnericanBy'LnkhtustLa are related to each other nore closely than either is tothe Ethiopian AlLuroldes, as is also indicated by the phyletic nearestneighbour relationships. A notable synaponorphy of Standeria andBz"Lnkfuivstia, with the fourth genus of the ALLuroididae, the nonoty?icBazrLejamLesonia Ljungstron, t971, (not cornputed) is the opening of theatrial prostates into the terninal chamber of the nale deferent appar-atus separately fron the vasa deferentia whereas in ALLtnoides the vasadeferentia discharge into the ental region of the tubular or buLbousatria. It has been suggested (Janieson, 1968b) that Standeria andBrinl<hu.rstia may be descended from a form with AlLuroides-type maLeterminalia. The phylogran is not inconsistent with this view but sug-gests that Alltuoides is paraphyletic rel.ative to the other two computedgenera. At present, however, no useful purpose would be served by split-ting the ALluroididae into two taxa of snprafanilial or higher rank;it is probable that autoaponorphies will be found to rmify the grotrp.

Accoding to Hennigian principles the Alluroidoidea nerit a rankcoordinate with the conbined Lunbricina and Moniligastrina, that issuperior to subordinal status. They are here, however, referred to anew suborder, the Alluroidina.

HapLotarLdae

The Haplotaxidae form the plesionorph sister-group of the conbinedsuborders Alluroidina, Moniligastrina and Lumbricina but are here referredto as the suborder Haplotaxina, a taxon proposed by Brinkhurst and Jamie-son (1971). I t is renarkable that the onLy apomorphies of the haplotaxidconputed (I/. uioLaeel*s) are short serninal vesicLes, egress of both pairsof vasa deferentia in a singl.e segment and absence of ovaries frorn XIII.H. gondLoides, the type-species (not conputed) lacks even these aponorphiesand is plesiomorph for aLl attributes employed in this study. This neansthat unless sone internal synaponorphy (autoapomorphy) be found for theHaplotaxidae it wi1l. have to be considered a paraphyletic grouping falselyconstructed on synplesionorphy of which the diagnostic one is locationof the nale pores one segnent behind the corresponding testes. In thisregard it is noteworthy that Gates (L972), for a different reason,recognized the possibi l i ty that the fanily is polyphyletic when he statedrrwhat now needs to be determined is whether and how nuch of the simplicityin haplotaxid sonatic structure is secondary and perhaps independentlyacquired in various areas separated from each other by wide oceanicdistancesrr. It is here considered that characters such as the unilayeredc1ite1lurn, large yolked eggs, the lateral l ines (the apparent honologuesof these l ines occurring in some polychaetes, Jamieson, 1968b) and octo-gonadial condition are unlikely to have been secondarily acquired and areindication that nany other features are genuinely plesionorph. Erectionof a separate family for progynous forns, including H. u'Lolaeeus, asopposed to the type-species on the grounds that the absence of sharedaponorphies indicates paraphyly would be of dubious value taxononically.The remarkable netagynous H. brinkhtusti recently discovered by Cook

222

B.G.M. JAMIESON

(1975) Lacks synaponorphic links with I/. gordLoi.des and E. uiolacetpbut siniLarly, is better included in the Haplotaxidae for the present.Brinkhurst and Jamieson (1971) have shown loss of the posterior pair ofovaries to be a haplotaxid trend but the discovery of a universal auto-aponorphy of the Haplotaxidae has yet to be nade.

0ther fani l ies

, The renaining groups of the Oligochaeta are the fanilies whichwere placed in a suborder Tubificina within the order of Hapl"otaxida byBrinkhurst and Janieson (1971). These are the aquatic nicrodriLe fan-il.ies typified by the Tubificidae and the very distinct fanily Enchy-traeidae. The additional oligochaete fanily, the Lunbriculidae, wasplaced in the nonotytrric order Lumbriculida. With the reinterpretationof the nonil igastr id testis-sacs and the relationship of then to the naleducts, the Lurnbriculidae becane the only faniLy with the prosopore con-dition (male pores opening in the sane segnent as the corresponding testes)and in view of their very few aponorphies it appears that if computed theywould have formed the plesionorph sister-group of the renainder of theOligochaeta. I t seens reasonabLe, therefore, to retain the Order Lunbri-cul ida.

The position of the Tubificina is uncertain. It now appears nostunlikely that if conputed they wouLd have intervened between the Haplo-taxidae and any of the remaining conputed forms. Yanaguchits concept ofa monotesticulate group for tubificids, naids, enchytraeids and siniJ.arlycharacterized nicrodriles seens to have considerable substance. Never-theless, the tubificid allangenent of gonads and nale terminaLia (onepair of testes and of firnnels in the segrnent preceding that containingnale pores, ovaries and oviducal funnels) is not derivable fron a lun-bricul id arrangement but, as suggested by Brinkhurst and Janieson (1971),is derivable frorn an octogonadial., protohaplotaxid condition. It appearsthat the Tubificidae, with the other nonotesticulate groups, conprise thesister-group of the opisthoporous oligochaetes (Haplotaxida) and an orderTubificida is here recognized for all nonotesticulate forms, the fornerTubificina. Lurnbriculids as the sister-group of the group HapLotaxida+Tubificida strictly rnerit supra-ordinal rank but are retained as an order.If , as Michael"sen (1928) and Yarnaguchi (1955) consider, the parasit icBranchiobdellidae and acanthobdeLl.id leeches are derived from protolun-briculids a strict application of Hennigian principles would demand thata monophyletic grotrping incl.uding these and forns currently regarded asoligochaetes were recognized.

Whether the tern 0l igochaeta could str ict ly be retained for thisgrouping would be questionabl-e. I f other leeches were considered toderive frorn a proto-acanthobdellid it would be necessary to include thenin the rnonophyletic grouping and the term Clitellata would rernain availablefor this group.

As the present application of Hennigrs principles to ol igochaetesystenatics has proved of considerable value it is to be hoped that asinilar study nay be conducted on what nay be terned the thecophorate(spermatheca bearing) anneLids, including the CliteI lata and the enig-natic Aeolosomatidae, and what nay be called the trochophorate annelids(Archiannelida and Polychaeta) .

223

PHYLOGENY OF TTM OLIGOCHAETA

Sunnary of the new classification

Full definitions are given only for new groupings.

0rder Lunbriculida

At least one pair of nale pores in the sane segnent as the corres-ponding nale f tnnels.Fanily Lr.rmbriculidae (As defined by Cook, t97L)

Order Tubif ic ida

Male pores 1 pair , in the segnent behind the single pair of testesand seminal funnels. 0varies 1 pair , in the segment behind the testes.Suborders Tubificina and Enchytraeina

Suborder Tubif ic ina

Spennathecal pores in segnents imrnediately in front of or behindthose bearing the nale pores, rarely in the same segnent.Superfanily Tubificoidea (As for the suborder)Famil ies : Tubif ic idae, Dorydri l idae, Naididae, 0pistocyst idae, andPhreodrilidae (As in Brinkhurst and Janieson, L97t)

Suborder Enchytraeina

Spermathecae anterior to the segnent bearing the reproductive organs,

separated by a gap of five segnents.Superfanily Enchytraeoidea (As for the suborder)nalnity Endlytraeidae (As in Brinkhurst and Jamieson, 1971)

Order Haplotaxida

Male pores 1 or 2 pairs in 2 successive segments or a single seg-

nent, at lLast one segment behind the posterior testes and seninal fun-

nels. With two pairs of testes and two pairs or more connonly one pair

of ovaries. I f testes reduced to one pair these are separated from the

ovaries by one or two segnents.Suborders Haplotaxina, Alluroidina, Moniligastrina and Lunbricina

Suborder Haplotaxina

Male pores l" or 2 pairs in 2 successive segments or a single seg-

ment, the last pair only one segment behind the poster ior of the two pairs

of testes and ftnnels. With two pairs of ovaries or only the anterior

pair or (H. br inkhurst i ) only the poster ior pair . Cl i tel lun one cel l

th ick.Superfanily Haplotaxoidea (As for the suborder)Fanily Haplotaxidae

As in Brinkhurst (1971) but including delet ion of the anter ior ovar-

ies and retention of those in XIII in HqLofurLs brLnkhursti.

224

B.G.M. JAMIESON

Suborder ALluroidina

Male pores 1 pair, in XIII; testes in X or X and XI; ovaries in XIII.Cli tel lum one celI thick.Superfamily Alluroidoidea (As for the suborder)farnify Alluroididae (As for the Al.luroidinae in Janieson, 1971d)Fanily Syngenodril.idae (As for the Syngenodrilinae in Jamieson' 1971d)

Suborder Moniligastrina

Testes t ot 2 pairs in testis-sacs each of which is suspended in theposterior septum of its segrnent and is detached fron the anterior septun;lnale pores 1 or 2 pairs, each pair at the posterior border of the segnentbehind the corresponding testis-sac. Cli tel lun one cell thick.Superfanil.y Moniligastroidea (As for the suborder)Fannily Moniligastridae (As for the suborder)

Suborder Lunibricina

Testes L ot 2 pairs, in X and XI or their homeotic equivalent; nalepores 1 pair or ("EopLoeltaeteLLd 2 pairs, 2 ̂ or nore segnents behind theposterior testes. nrhere there ate 2 pairs of testes the 2 vasa deferen-iia of a side uniting at the naLe pores (exceptiona1.ly, HoploehaeteLla'remaining separate); ovaries a pair in XIII or i ts honeotic equival.ent,rarely wittr a second pair in the preceding segnent. CLitel lum nult i-Layered.SuperfaniLies: Lrlribricoidea, Biwadriloidea, Sparganophil.oidea, Alnoideaand MegascoLecoidea.Superfani ly Biwadriloidea-

Lateral Lines present between the ventral and dorsal setal. couples.Body grooved dorsalJ.y but cross section not quadrangular. Anus dorsal,subterrninal. Dorsal pores absent. Testes two pairs, in X and XI. Ovar-ies lobuLated, in XIII. Calciferous glands, typhlosoles and gizzard-l ikethickening of the aLimentary nusculature absent; intestinal origin indef-inite, in the region of XI-XIII. Supra-oesophageal blood vessel present.

Subneural vessel absent. Lateroparietal vessels 2 on each side, oneoriginating fron the dorsal, the other fron the suboesophageal vessel.Lasi heart i in Xt. Male pores 1 pair, in XIII, innediately precLiteLlar.Vasa deferentia embedded in the bodywall. Lobul.ar nasses of prostaticglands discharging near the nale pores together with a pair of prostate-Iike setal glands. Nephridia holonephridia; absent from the forebody;tacking terminaL bladder, caecum or sphincter; all' exonephric. Sperna-thecae absent. cl i tel lum extensive, mult i layered, aruruLar.Fanily Biwadril.idae (Definition as for the Biwadrilinae Janieson 1971d)Superfami ly SParganoPhiloidea'

Latetal lines absent. Anus dorsal but body not quadrangul'ar. Dor-sal pores absent or sporadical ly present intraspecif ical ly. Setae sigmoid,sinpie pointed, 8 per segrnent. Testes two pairs, in X and XI. Ovaries

1 pi ir,-single-str inged, in XIII. Calciferous glands, gizzards typhlo-soie and mulcular thickening of the intestine absent. Intestine conmen-cing in IX. Supra-oesophageal and subneural vessel.s absent. Lateropa-rieial vessels 2 on each side, originating fron the dorsal and ventral

225


vessels. Last hearts in XI. Male pores intracl i tel lar anteriorly in XIX.Prostates absent but prostate-like glands nay be present in a few to nanysegments. Vasa deferentia deeply ernbedded in the body wall. Spernathecaeintracoelomic, pretesticular, paired ol nult iple.Farnily Sparganophilidae (As for the Sparganophilinae Jamieson, 1971d).Superfamily Alnoidea

Lateral lines absent. Body quadrangular in cross section; dorsalgroove (always?) present; anus dorsal or dorsoterninal; dorsal pores ab-lettt . detae-sigrnoid, simple pointed, 8 per segnent. Testes 2 pairs or(LutodriLus) 10 pairs in XII-XXI by interpolation of 10 segnents, of whichthe last 8 are testicular, anterior to the normal 2 pairs of X and XI.Ovaries 1 pair, with nany egg-strings' or (CrLodrLlus) not forming egg-str ings, in XII I or (1 species) persist ing in XII a lso. Digest ive systennot usually possessing oesophageal gizzard(s) but thickening of the anteriorintestinal musculature usua1. Intestine comnencing in or far behind )ff(to XXXVIII) or (Lutodri l idae) di lat ing gradually in posterior gonadal.segments; dorsal t lphlosole ptesent; calciferous gLands absent. Vascularsysten with dorsal, ventral and often supra-oesophageal and subneuralvessels. I f a supra-oesophageal is present, sone hearts (al.ways?) latero-oesophageal. Last hearts in XI or less conunonLy XII or XIII or (LutodrLLus)in XI-XXI. Male poaes 1 pair, in or behind XV, intra- or ante-cl i tel lar.Longitudinal. tubercula pubertatis absent or indistinctly developed orrepresented by al i form ridges. Prostate 9Lands or prostate-l. ike 9Landsrarely present; prostate glands if present lacking ectaL nuscular duct.Vasa deferentia deeply enbedded in the body wall, sometines dischargingthrough glandul-ar bursae. Spernathecae adiverticulate, post-testicular,rarely extending into or anterior to the testis segnents; largely intra-parietal; never with clearly differentiated intracoelomic duct; usuall.ylnultiple in each intersegnent occupied; rarely absent. Nephridia absentfron 10 or nore anterior segnents; all exonephric holonephridia, sone-tines vesiculate but never with diverticuLun; terminal sphincter absent.Clitellurn multiLayered; aru:uJ.ar. Cocoon elongate, fusiform, with several.to nany enbryos.Famil ies: Alnidae, CriodriLidae (Definit ions as for the Alminae andCriodri l inae; Janieson, 1971d) and Lutodri l idae McMahan' t976,Superfamily Lr-unbricoidea. Ernend.

Lateral l ines absent. Body approxinately circular in cross sectionwith posterior terninal anus; exceptionally (EisenielLd quadrangular withdorsoterminal anus. Dorsal pores present or absent. Setae signoid, sinplepointed (exceptionally bifid?), 8 or (Pertseolen) numerous per segnent.Testes two pairs, in X and XI (or one pair in one of these). Ovaries1 pair, fan shaped or single-str inged, in XIII or (1 species) persist ingin XII also. Digestive systen usually with oesophageal gLzzatd(s) (pre-gizzatds) less frequently with thickening of the anterior intestinal giz-zard (post gizzard) or with pre- and post-gizzards. Intestine usuallycommencing in XV. Vascular system with dorsal, ventral and often withsupra-oesophageal and subneural vessels. I f a supra-oesophageal ispresent, some hearts (always?) latero-oesophageal. Last hearts usuallyin XI, sometimes more posterior. Male pores 1 pair, rarely 2pairs, inXV to many segments more posteriorly; alnost always anterior to the nid-cl i tel lun, often precl i tel lar, exceptionally (Opistodtdlus) postcl i tel lar,opening directly or fron glandular or muscular bursae but prostate glands

226

B.G.M. JAMIESON

rarely present and never (?) with ectal nuscular ducts; euprostates unknown.Vasa deferentia frequently deeply enbedded in the body waLl.. Spernathecaeadiverticulater V€rI variable in number and location, pre- or post-testicu-lar, paired or mult iple; or absent. Excretory systen hoLonephric or(TrLtogewta) nephridia duplicated; sometimes with anterior enteronephry,vesicul,ate or avesiculate. Cli tel" l .un nult i layeredrusually saddl.e-shaped,often associated with longitudinal tubercul.a pubertatis.Fanil.ies: Lurnbricidae, Horrnogastridae, (both redefined bel.ow), Microchaetidae,(Definit ion as for the Microchaetini Janieson, 1971d), Kynotidae (Defin-i t ion as for the Kynotinae Jamieson, 1971d), GlossoscoLecidae (Definit ionas for the Glossoscolecini Janieson, 1971d) and Komarekionidae Gates, L974,Fanily Lunbricidae Claus L876. Emend.

Prostonium prolobous to tanylobous, never? zygolobous. Forn cylin-drical or rarely posteriorl.y quadrangular or trapezoidal in cross section.Pignent present or absent. Setae in 8 l.ongitudinal rows comnencing on II,signoid, sinple pointed, often cicatr iced distal ly, cLosely or widelypaired or unpaired. Some ventral. setae often on genital tr.unescences andnodified as genital setae each of which (always?) bears 4 longitudinalgrooves. Nephropores on each side in a single series or aLternating.Clitel. lun usual. ly saddle-shaped, beginning between the 17th and 52nd seg-rnents and occt4rying 4 to 32 segrnents. Tubercula pubertatis situated inthe clitellar region and shorter than to longer than the clitellun, forminga pair of continuous str ips or r idges; or series of pairs of isolatedpapil lae or sucker-l . ike tubercles. MaIe pores intrasegnental., antecl i-teLlar, alnost always on XV, sonetimes on XIII, rarely on XI or XII,often in slits in glandular tumescences each of which nay correspondinternally with a short glandular atriun projecting into the coel.orn butprostates with muscular ducts absent. Fenale pores on XIV. Spernathecalpores usualLy in front of the ovaries, never on the cl i tel lum; 2-B pairsin intersegnents 5/6-L9/20, varying in position fron setal lines a tothe niddorsal Line but usually in 9/10 and L0/lL in cd l ines; sonetinesnultiple in each segment occupied; sonetimes absent. DorsaL pores pres-ent in nedian singl-e file or bilateral pores present in Line with thedorsal. setal couPles.

0esophagus l.acking gizzatds but with calciferous glands in X, XI-XIII, XIV, )ff (al.ways?) comprising longitudinaL chanbers that open attheir anterior ends into the oesophageal. lunen which may or nay not formextramural pouches or diverticula. The intestine beginning in XV or(DLpozodrLlts) XIV or XVI with a I'croprrfoll.owed by a gizzatd. Thegizzard occupying 1 to 3 segments, usually in XVII and XVIII, neverbehind XX. Typhlosole (always?) present. Intestinal caeca and supra-intestinal glands absent. Dorsal, ventral and subneural blood vesselswelL developed; the subneural adherent to the nerve cord. Hearts 5-7pairs, of which the last pair is in XI or (0, oofu.edrd 3 or 4 pairswith the last in IX or X. Supra-oesophageal vessel absent. Latero-oesophageal vessels rnedian to the hearts and passing to the dorsalvessel in the region of X-XI. Lateroparietals absent. Nephridiaholonephridia, vesiculate, the bladder sonetimes with divert iculum.Nephridia sometines discharging into longitudinal ureters. Testes inX and XI or, tately, in only one of these, free or in suboesophageal orperi-oesophageal testis-sacs (the tatter type of sac enclosing the heartsand seminaL vesicles of X and XI). Spenn ducts apparentLy never concealed

227


in the body walI musculature; sometines expanded or coiled behind thesperm funnels to forn epididyrnes. Seninal vesicles 4 pairs, in IX-XII:3 pairs, in IX, XI and XII, or 2 pairs, in XI and XII. Ovaries in XIII,always (?) with a single terminal string of oocytes. Oocytes maturingin ovisacs which project into XIV fron its anterior wall. Spermathecaeadiverticulate though sometines appearing diverticuLate owing to cons-tr ict ion by septa; sessi le or pedurculate but Lacking dist inct duct.Subfanilies: Lr-unbricinae and DiporodrilinaeSubfanily Lunbricinae

Definition as for the fanily with dorsal pores in rnedian single fi1,e,lacking paired, bi lateral. coelonic pores.Subfami ly Diporodri linae

Differing from the Lunbricinae in lacking dorsal pores and possessingbilateral coelonic pores. Prostomiurn epil .obous. Form cyl indrical. Lack-ing cutaneous pignentation. Sculpturing of sonatic and genital setaenoi recorded. Nephropores on each side in a straight l ine. Cli tel lunin the region XXI-XXXIII. Tubercula pubertatis shorter than the cl i tel-Ium. Mal- pores on XV. Spermathecal pores in the region 8/9'LI/L2.Hearts in VI, VII-XI but vascul.ar systern othervrise undescribed. Cropwithin the region of XIV-XVII; gizzatd in two of segnents XVII-XX. Pos-terior nephridia retaining segnental. pores but interconnected on eachside by a longitudinal ureter. Holandric; testis-sacs absent. Seninal"vesicles in XI and XII. Spernathecae intraparietal or intracoel 'onic,sessi le.Fanily Hormogastridae Michael'sen, L928. Enend.

-Seta" 8 per segment, closely paired; genital setae Longitudinal" ly

grooved. Dorial pores absent. Male pores intracLitel lar, anteriorly6" portutiorLy on the clitellum, varying from tZ XV to 15/16 or (Ailos-

col-ecinae) at 1< XXII; on or anterior to tubercuLa pubertatis if- theseare present; such tubercula groove-l ike. Fenale pores between L XIVanaiq/l$, varying fron ventral to a to the region of c Lines. Last

hearts in XI. Oesophageal glzzards (progizzards) well developed. In-testinal gizzards (postgizzards) rudinentary or absent. Calciferousglands present or abseni. Holonephric; nephridia vesiculate and (always?)

witn aivert iculun. Holandric; testis-sacs absent; two pairs of shortseminal vesicles, in XI and XII. Prostates absent or (Ailoscolecinae)

sessile prostate-l ike glandular nasses associated with the tuberculapubertatis in the vicinity of the nale pores. Metagynous; ovaries single-str inged(f). Spermathecae adivert iculate, paired or nrult iple, always with

,or" In it e g"nltal segnents but sometines extending anterior to then.subfamil ies Hornogastrinae, vignysinae and AiloscolecinaeHonnogastrinae Bouch6, 1970

Hornogastridae with 3 dist inct gizzards, in VI, vII and VIII.Rudinentary inteut:rnaL gizzards present or absent. No blood vesicleson the ventroparietal vessels. Body wal1 not especial ly thickened.With or without calciferous glands. Fenale pores in the region ofb lines. Male pores varying fron %XV to 15/16, on or anterior to the

tubercula pubertatis if these are present. Spernathecal pores paired

or nul t iPle.Vignysinae Bouch6, L970

Hornogastridae with 2 rmequal gizza'rds in VI and 6/7 and with sub-parietal blood vesicles. Intestinal gizzard present. Calciferous glands

228

B.G.M. JAMIESON

absent. Male Pores posterior ' in XV, well anterior to the tuberculapubertatis which are groove Like. Fenale Pores ventral to a lines.Spermathecal pores Paired.Ai1oscolecinae Bouch6, 1969

Male pores intraclitellar, discharging on tubercula pubertatis

at %XXII; tubercula groove like. Spernathecal pores imurediately an-

terior to the testes. two ptogLzzatds, in VI-VII and VIII-IX. Intes-

tinal gizzard, absent. Calciferous glands present. Prostate glands

associited with the tubercula pubertatis in the vicinity of the nalepores. FenaLe pores shortly above b lines. spermathecaL pores paired.

Superfami ly Megas co1 ecoidea- LateraL lines absent. Body approxinately circular in cross sec-

tion with posterior terminal anus or if not circular not quadrangular

and not wiih dorsoterninal. anus. Dorsal pores usually Present. Setae

rigroia, simpLe-pointed, 8 P9r segnent or.perichaetine. Testes two pairs,

in-X and XI ior ih"i" honeotic equivalent) or one pair in one of these.

Ovaries f pair with several to nany egg str ings in XIII or (1 species)persist ing i t t Xff also. Digestive systen often with oesophageal ' giz-'zard(s), iarely with thickening of the anterior intestinal musculature.

fnteitLne begiirning in XII or XIII or, usually, more posteriorly. Vas-

",rf." systen with dorsal, ventral and often with supra-oesophageal and

subneurlL vessels. If a suPra-oesophageaf is present' some hearts(al.ways?) latero-oesophageal.. Last hearts in or, usual ' ly, behind xI.

f'1"f" iro""s 1 pair or rarety (Hoploelnetella part) two pairs, in )WII-

XX, c-omnonfy -lWff

but predominantly )ffIII, shortl.y behind or anterior to

ttre posteribr end of tire cl.ite1l.urn, rarely anterior to the nidcl'iteLl'un.

Usuaity with prostates, at oT in the vicinity of the nale pores, the

glanduiar part of which nay have nuscular walls and receives the vasa

ieferentii entall.y (euprostates) or with insignificant nusculari zation

oe tn" glandular part but usually with nuscular ectal duct and with

the vasi deferentia entering the duct ectal to, or rarely within, the

lianaufar portion (netaprosiates). Metaprostates tubuLar, with sinple

fnternal duct, or tubuLoracemose, with lateral}y branched or diverti-

culate lunen, or racenose, with ranifying internal ducts. Vasa defe-

rentia not deeply enbedded in the body wall. Spernatheca(e) paired,

sonetines ,-p"i"La or mrl.tiple, diverticulate or less conrnonly adiver-

ticuLate, raiely absent, the diverticula usually sperm-storing; predon-

inant!.y with spermathecae pretesticular but (EudriJ'idae) showing a

tendenty to beione post-testiculate or even posterior to the male pores;

if post-testicuLar ?requently corununicating with the oviducal apparatus'

Excretory systen holonephric and/or neronephric, exonephric and/or

enteronephric, sonetines vesiculate. CLitel lun rnult i layered; annular

rarely s-addte-shaped; longitudinal tubercula pubertatis rare'

Fanilies : Megascolecidae, Eudrilidae and 6cnerodrilidaeFamiLy Megas-olecidae (Ai in Janieson, !97La, but excluding the Ocnero-

dri l inae) .Subfarnily Megascolecinae (As in Janieson, 1971c) 'Subfanily AcinthodriLinae (As in Janieson, 1971b) 'Fanily Eudri l idae (As in Gates, t972) 'Subfamilies Eudrilinae and Pareudrilinae (As in Gates

' 1972) '

f .r i fy gcnerodri l idae (As for the 0cnerodri l inae in Janieson, 1971b).

229


Subfanily Ocnerodrilinae (As for the tribe Ocnerodrilini in Janieson,1e71b).Subfamily Malabarinae (As for the tribe Malabarini in Jamieson, 1971b).

Acknowledgenents

The author is indebted to Mr. C. Andrews, Departnent of ComputerScience, University of Queensland, for modifying his CLUSTR programfor Hennigian anal.ysis and for help freely given and Mr. C. McGovern,Prentice Conputer Centre of the University, is thar*ed for sinilar aid.Dr. M. Dale, C.S.I .R.0. , is a lso thanked for independent val idat ion ofthe conputations. Special thanks are due to ny research assistant,Mr. John Bennett, for his careful preparation of the diagrams fron con-puter printouts and for contributing to soLution of computational prob-lens. This work wouLd not have been possible without grants fron theAustralian Research Grants Comnittee and the Australian Biol.ogicalSurvey which a"e gratefuJ.ly acknowl.edged. Mrs. V. Maume is thankedfor her careful typing.

References

Andrews, C.J. L976, Users manual for CLUSTR, a comprehensive suite fornumericaL cl.assification. Pilot nanual of the Departnent ofConputer Science, University of Queensland.

8a11, I.R. 1974, A contribution to the phylogeny and biogeography of thefreshwater tr icl.ads (PLatyhel.rninthes :Turbellaria). In: Riser, N.W.and Morse, M.P. (eds.) . Biol .ogy of the Turbel l .ar ia. McGraw Hi l l ,New York 15: 339-401.

Beddard, F.E. 1895. A nonograph of the order 0l igochaeta. Oxford,Claredon Press.

1901. Oligochaeta. In Canbridge Natural History 2.f f iT.-rggo.Anattempttoc] .assi fyearthworns.Q.J1nicrosc.

Sci . 3L: 201-315.Bouchd, M.B. 1969. AiLoscolex lacteospumosus, n. gen., n. sp. Un ver de

terre aux caractbres rnorphologiques et biologiques renarquables(0l igochaeta, Ai loscolecidae, nov. fam.). Rev. Ecol . Biol . So1.6,42 525-531.

1970. Renarques sur queLques Lumbricina de France et con-

@es de 1a decouverte des nouveaux taxons Vignysinae (Subfan.

nov.) et Diporodr i l idae (Fam. nov.) . Pedobiologia, 10: 246'256.L972. Lonbric iens de France. Inst i tut Nat ional de la

---.ffierche Agrononique, Paris.Brinkhurst, R.O. and Janieson, B.G.M. L971. The Aquat ic Ol igochaeta

of the World. Ol iver and Boyd. Edinburgh.Brundin, L. 1965. On the real nature of t ransantarct ic relat ionships.

Evolut ion, 19: 496-5051966. Transantarctic reLationships and their significance

--ffiaenced by chironornid rnidges. K. Svenska Vetenskakad. Handl.1-1 : L-472.

Canin, J.H. and Sokal, R.R. 1965. A nethod for deducing branchingsequences in phylogeny. Evolut ion 19: 3LL-326.

Col less, D.H. t967. The phylogenet ic fa l lacy. Syst. Zool . 16: 289'295,Cook, D.G. L97L. In: Br inkhurst , R.0. and Janieson, B.G.M., The Aquat ic

0Ligochaeta of the World. Ol iver and Boyd, Edinburgh.

230

B.G.M. JAMIESON

Cook, D.G. 1975. Cave-dwell ing aquatic 0l igochaeta (Annelida) fron theeastern Uni ted States. Trans. Am. Microsc. Soc. 94,12 24-37.

Crowson, R.A. 1970, Classif ication and biolog;y. Atherton Press, N.Y.and Heinenann Educational Books, London.

Gates, G.E. L942. Notes on various peregrine earthworns. Bull . Mus. comp.ZooL. Harv. 89,3: 6L-I44,

1945. 0n the Oligochaete genus Syngenodrilus and its taxo-f f i re] 'at ionships.J.Wash.Acad.Sci .35:593-596.

1959. On a taxononic puzzLe and the classification of the---G-worms. 8u11. luftrs. conp. zool. Harv. t?t: 229-26t.

L972. Burrnese earthworms. Trans. Atn. phil . Soc. (n. ser)@t-s26.

t974, Contributions on North Anerican earthworms (Annelida)--- f f ionanewspeciesofearthworminasouthernport ionofthe

United States. Bul1. Tal1. Tinbers Res. Stn. 15: 1-13.L976, 0n earthworm ovaries and their importance in negadrile

ff iat ics. Megadri logica 2,L22 L-4.Ha4rer, C.W. 1976. Phylogenetic inferences in paLaeontology. J1. Paleont.

50,1: 180-193.Hennig, W. 1966. Phylogenetic Systenatics. University of I l l inois Press.

Urbana.Horst, R. 1889. 0ver eene nieuwe soort order de Lumbricinen door Prof.

Max. Weber uit nederL. IndiH medegebracht. Tidjdshr. ned. dierk.Vereen. 2: 77,

Hull, D.L. L970. Contemporaly systematic philosophies. Ann. Rev. Ecol.Syst . 1: 19-54.

Janieson, B.G.M. 1957. Sone species of Pygnaeodri l .us (Oligochaeta) fromEast Africa. Ann. Mag. nat. Hist. (12) 10: 449-468.

1963. A revision of the genus GordiodriLus Beddard@:Megasco1ecidae).Bu]. ] ' .Br i t .Mus.(Nat.Hist . )Zool .

9,8: 297-323.1965. Tho Ocnerodril.inae (Megascolecidae, 0ligochaeta)

--forn East Africa: Nematogenia lacuun (Beddard 1895) and PygnaeodriLusnont iskenyae sp. n. Ann. Mag. nat. Hist . (15) 8: 95-107.

t967. New records of Ocnerodrilidae and Lumbricidaeff i f r icanr ivers.Zoo1ogicaAfr icana3,1:59-86.

@'?:n,LTx,iil;:";."f":i'fl*r'lI'5Yi:il:'*tidae:1968b. A taxonometric investigation of the Al.luroididae

@). J. ZooL. Lond. (196s) 155: s5-86.

@'ii?;t':::":ffi::":''ffi:lii.i''!}:li:*llHjffT;llf:",lel isn with the genus Sttrhhnannia. . I . nat. Hist . 3: 41-51.

1970a. A revision of the Australian earthworm genusff iawithdescr ipt ionsoftwoneWgenera.J.ZooI.Lond.

L62: 99-144.1970b. Two new syrnpatric species of the earthwofln genus

@gasco1ecidae:01igochaeta)fronQueens1and.Proc.R.Soc. Qd. 82,3: 35-46.

1970c. A taxonomic revision of the Ol igochaete genusEukerr ia Michaelsen, 1935 (Ocnerodri l inae:Megascolecidae) . 8u11.Br. lv t t rs. nat . Hist . (Zoo1.) 20,5: L3L-L72,

23L

PTryLOGENY OF THE. OLIGOCHAETA

Janieson, B.G.M. L97La. A review of the Megascolecoid earthworm genera(0l igochaeta) of Austral ia. Part 1- Reclassi f icat ion and checkl istof the megascolecoid genera of the world. Proc. R. Soc. Qd.82,6:75-86.

1971b. A review of the Megascolecoid earthworm genera(Oligochaeta) of Austral ia. Part I I - The subfamiLies 0cnerodri l inaeand Acanthodr i l inae. Proc. R. Soc. Qd. 82,8:95-108.

1971c. A review of the Megascolecoid earthworn genera@)ofAustra1ia.Part I I I .Thesubfani1yMegasco1ecinae.

Men. Qd. Mus. 16,1: 69-L02.L97Ld. In: Br inkhurst , R.0. and Jamieson, B.G.M., The

@ochaetaoftheWor] .d.01iverandBoyd.Edinburgh.L97Le. Earthworrns (Megascolecidae: Oligochaeta) fron

@ral iaandtheirzoogeography.J.ZooL.Lond.165:47L-504,L972a. A numerical study of the Australian earthwor:n

@odri1usF1etcherandTrinephrusBeddard,withanewgenus (Megascolecidae:Ol igochaeta). Zool . J. L inn. Soc. 5t ,2: 147-L75.

t972b. Redescription of the type species of the earthworngenus Notoscolex (Megascolecidae:0l igochaeta). J. nat . Hist .6: 515-519.

L974a. The indigenous earthworms (0Ligochaeta:[email protected]] ' .Br.Mus.nat.Hist .Zoo1' .26'3. .2o3.328.

1974b. The earthworrns (Oligochaeta:Megascol.ecidae) off f i t ia. Proc. R. Soc. S.A. 98,2: 79-L12.

1974c, Generic type species and other Megascolecidae@1igochaeta) intheMuseunofSystemat icZoo1ogy,t ln iv-

ersi ty of Tur in. BolL. Mus. ZooL, Univ. Tor ino 8: 57-88.7975. The genus Digaster (Megascolecidae:Oligochaeta)

in Queensland. Mem. Qld. Mus. 17,22 267-292.Jamieson, B.G.M. and Nash, W. t976. The f irst record of Plutel lus (Megas-

coLecidae:Oligochaeta) in Queensland, P. incomnodus and P. ravenispp. nov. Proc. R. Soc. Qld. 872 45'52.

Janieson, B.G.M. L976, A re-examination of the type species of DiporochaetaBeddard, 1890 (Megascolecidae:Ol igochaeta). J l . R. Soc. N.Z. 6,1:37 -44 .

t977a. 0n the phylogeny of the Moni l igastr idae with anew species of Moni l igaster (Ol igochaeta:Annel ida) . Evolut ionaryTheory 2: 95-114.

L977b. Prelininary descriptions of Indian earthworms

@dae:o1igochaeta)fronthePa1niHi11s.Bu11.Mus.Natn.Hist . Nat.r , Paris , Zoologie 31.3,Se'r450 z 477-502.

Ljungstron, P.O. 1971. New species of aquatic earthworrns from Malawiand South Afr ica. Rev. Ecol . Bio1. So1. 8,22 313-325.

L974, On two species of EudriLidae from Malawi and

@esurrect ionofthefarni ] .ya1rnidae(O1igochaeta).Rev.Ecol . Biol . So1. L1,1: t4 l -L48

Marx, H. and Rabb, G.B. 1970. Character analysis: an imperical approachappl ied to advanced snakes, J. ZooL. London, 161: 525-548.

McMahan, M.L. 1976. Prel i rninary notes on a new negadri le species' genusand fanily frorn the southeastern United States. Megadrilogica,2 ,11: 6- 8.

232

B.G.M. JAMIESON

Meyer, A. L929. Zur Segrnentierungsanalyse und Stanngeschichte des OJ.igo-chBten. ZooL, Anz. 86: 1-16.

Michaelsen, W. 1920. OligochBten von WestLichen Vorderindien und ihreBeziehungen zur OligochBtenfauna von Madagaskar und Seychellen.Mitt. naturh. Mus. Hamb. 38: 27-68.

t92L. Zur Stanmresgeschichte und Systernatik der Oligochften,f f i .erederLurnbr icu] . iden.Arch.Naturgesch.86,8:130.141.

L928. OJ.igochaeta. In: Kukenthal, W. and Knmbach, T.,Handbuch der Zoologie 2,8: 1-116.

1930. Ib id. Nachtrag 1: 117-118.L932, Ib id. Nachtrag 2: 118.

NAaseT, anA Nonura, E. L937. 0n the Japanese aquatic OligochaetaCriodr iLus niyashi tae, n. sp. Sci . Rep. Tohoku. Imp. Univ.4:36L-402.

Onodeo, P. L948. Ol. igocheti delLa Canpania. I I . Morfologia e biologiadei terr icol. i : Microscolex dubius (Fletch) (Megascolecidae, Acan-thodrilinae) ; Hormogaster redii Rosa (Glossoscolecidae, Hormogas-tr inae). Armuar. R. Mus. ZooL. R. Univ. Napol i 2: 1-11.

1956. Sistenatica e distr ibuzione geografica degli Hormogas-rce(01igochet i ) .Archo.Bot.Biogeog.I ta1.32,t ,4:L59-L77.

1958. La reserve naturel le integrale du nont Ninba. I. 0l. igo-rutes. Mem. Inst . f r . Afr . noire 53: 9-109.Pickford, G.E. 1945. Addit ional. observations on the 0l igochaete Genus

Syngenodr i lus. J. Wash. Acad. Sci . 35: 397-399.1975. Contributions to a study of South African Microchae-

TinrAnne1ida:0] . igochaeta).Trans.Conn.Acad.Ass.Sci .46zt3-76.Righi, G. t97L. Sobre a fanil ia Glossoscolecidae (01,igochaeta) no Brasi l .

Arq. ZooL, S. Paulo 20rL: 1-95.L972. Topography of the circulatory systen of Glossoscol.ex

-- paultstus Michaelsen 1926 (Oligochaeta Gl.ossoscolecidae). Mon. ZooL.i ta l . . 6: 19-56.

Ross, H.H. 1974. Biological Systenatics. Addison-WesLey, Reading' Mass.Sirns, R.W. 1966. The classif ication of the Megascolecoid earthworms:

an investigation of Oligochaete systenatics by Computer technique.Proc. L inn. Soc. Lond. L77,2: L25'141.

L967 . Earthworns (Acanthodrilidae and Eudrilidae :Oligochaeta)%nitanbia. Bul l . Br. Mus. nat. Hist . 16,1: I '43.

L973. Lumbricus terrestr is Linnaeus, 1758 (Annelida, Oligo-f f i .a) :designat ionofaneotypeinaccordancewithaccustoned

usage. Problens arising fron nisidentif ication of the species bySavigny (L822 61826) Z.N. (S.) 272. Bul . l . ZooL. Nonencl . 50'1:27-3L.

Smith, F. and Green, B.R. 1919. Descript ions of New African earthworns,incLuding a nehr genus of Monil igastr idae. Proc. U.S. Natn. Mus.55:145-166.

Sneath, P.H. and Sokal, R.R. L973. Nunerical Taxonomy. Freeman, SanFrancisco.

Sokal, R.R. 1966. Nurnerical taxonony. Scientif ic Anerican 215,62 106-117.Sokal, R.R. and Sneath, P.H. 1963. Principles of Nunerical Taxonony.

Freenan, San Francisco.Sporne, K. 1948. Correlation and classif ication in the Dicotyledons.

Proc. L inn. Soc. Lond.160:40-58.

233

PHYLOGENY OF THE OLIGOGIAETA

Stephenson, J. 1930. T?re Oligochaeta. Oxford. CLarendon Press.Yarnaguchi, K. 1955. Studies on the aquatic Ol"igochaeta of Japan VI.

A systematic report, with some remarks on the classification andphylogeny of the Oligochaeta. J. Fac. Sci. Hokkaido l}niv. Ser. VI,Zoology, I I , 2z 277-342.

phylogenetic and phenetic systematics of the …

Documents