Recordsofthe WesternAustralianMuseum 23: 133-144 (2006). Osteology of the first dorsal fin in two terapontid fishes, Leiopotherapon unicolor (Gunther, 1859) and Amniataba caudavittata (Richardson, 1845), from Western Australia: evidence for hybridisation? DavidL. Morganand HowardS. Gill CentreforFishandFisheriesResearch,MurdochUniversity,SouthStreet,Murdoch, WesternAustralia6150,Australia;Email: [email protected] Abstract - Osteological characters, such as number of supraneural bones anteriorto firstneural spine, numberofspineson first dorsal pterygiophore, position of insertion of first proximal dorsal pterygiophore and number of anterior proximal dorsal pterygiophores insertingbetween successiveneural spines, in conjunction with morphological characters, were used to provide evidence of natural hybridisation between two species of the Terapontidae: thefreshwater Leiopotherapon unicolorand themarine/estuarineAmniataba caudavittata. INTRODUCTION Western Australia, throughout northern Australia The Terapontidae (commonly called trumpeters and southern New Guinea to the east coast of or grunters) consists of small to medium-sized northern Queensland (Vari 1978; Allen etal. 2002). fishes represented by approximately 46 species WhilstA. caudavittataisconsideredtobeprimarily from 16 genera in marine and freshwaters of the marine (Vari 1978), in south-western Australia it is Indo-Pacific region (Vari 1978; Nelson 1994; Allen essentially restricted to estuaries (Potter etal. 1994; etal. 2002). Of these, about 33 are restricted to the Wise et al. 1994). It is also able to tolerate both freshwaters of New Guinea and Australia where freshwater and hypersaline conditions (Hutchins they are often of economic and/or recreational andSwainston 1986;MorganandGill2004). importance (Allen et al. 2002). The 16 genera Leiopotherapon unicolorandA. caudavittataboth mm assigned to the group are separated on the basis of attain total lengths of ca 300 and are the following features: pigmentation; extrinsic superficially similar, with considerable overlap swimbladder muscle; swimbladder and intestinal occurring in many of the characteristics often used pattern; height of dorsal and anal fin sheath; and to differentiatebetween species of fish, e.g., dorsal- osteological characteristics of the posttemporal, fin spines (XI-XIII in L. unicolor cf. XII-XIII in A. tabular,dentary, maxilla, premaxilla,spinousdorsal caudavittata); dorsal-fin rays (9-12 cf. 8-10); anal-fin fin and vertebral column (Vari 1978). The rays (III, 7-10 cf. Ill, 8-9); pectoral-fin rays (15-16cf. Australian terapontid fauna consists of 25 species 13-17); pelvic-fin rays (both 1-5); lateral line scales which are restricted to freshwaters and a further (45-57cf. 46-54); scales above lateral line (both 7-9); eight species that are considered predominantly scales below lateral line (16-20 cf. 17-19); caudal marineorestuarine(Vari 1978;Allen etal. 2002). scales (3-6 cf. 4-6); and predorsal scales to occiput Two of the most common and widespread (15-20 cf. 14-17) (Vari 1978). Notwithstanding the terapontids in Australia are the Spangled Perch above similarities in these species, L. unicolor is Leiopotherapon unicolor (Gunther, 1859) and the readily distinguished from A. caudavittata as the Yellowtail Trumpeter Amniataba caudavittata fins of the former species are a uniform pale, (Richardson, 1845). The former is the most silvery-grey colour, whereas those of the latter widespread freshwater fish species in Australia, speciesareyellow and, inthecaseofthe2nd dorsal- occurring in major rivers, isolated drainages and fin and caudal-fin, bear prominent black bands ponds throughout the northern two thirds of (Vari 1978). Furthermore, in L. unicolorthe row of Australia, includingmostriversystemsnorthofthe spots running from the eye to the upper jaw and Murchison River in Western Australia, the the row below the eye from the preopercle to the Northern Territory, Queensland, the Lake Eyre/ upper jaw are represented by distinct spots, Bulloo drainage systems and those rivers north of whereas in smaller specimens of A. caudavittata the Murray-Darling River in New South Wales thesespotscoalesce to form two distinctbands (see (Vari 1978; Allen et al. 2002). Amniataba Figure 1 and photographs on pages 227 and 241 in caudavittata occurs from the lower west coast of Allenetal. 2002). 134 D.L.Morgan,H.S.Gill Figure1 A, Amniataba caudavittata. B, Leiopotherapon unicolor and C and D, intermediate forms (note bars on caudalfinsandroundedsnout). Hybridisationof L. unicolorandA. caudavittata 135 In addition to the clear distinction in coloration Hybridisation between fish species is not between L. unicolorand A. caudavittata, members uncommon and Schwartz (1972, 1981) compiled of the genus Leiopotherapon can be differentiated almost 4000 references reporting either natural or from those of Amniataba using the following artificial hybridisation between various species. osteological criteria: one supraneural bone in front Historically, themain methods ofdetectinghybrids of the first neural spine in Leiopotherapon have been through the comparison of compared to two in Amniataba; one spine on first morphological characters (e.g., morphometric and dorsal pterygiophore cf. two; the insertion of the meristic data), with the assumption that the hybrid first proximal dorsal pterygiophore between the has characters that are intermediate between the second and third neural spines cf. its insertion parent species (Campton 1987, 1991). In this paper between the first and second; two of the anterior wedescribetheosteologyofthefirstdorsal finand, proximal dorsal pterygiophores inserting between bycomparison with thedescriptions and diagnoses successive neural spines cf. one (Figure 2) (Vari of the two genera provided by Vari (1978), provide 1978). evidence of natural hybridisation in the Duringa recent surveyofthe freshwaterfishes of Terapontidae. In addition, we provide evidence to the Pilbara (Morgan and Gill 2004) L. unicolorand suggestthatthehybridsarereproductivelyviable. A. caudavittata were often observed schooling together up to 300 km inland in the Murchison River. On closer examination, it became apparent MATERIALSANDMETHODS that many individuals from these mixed schools possessed combinations of coloration and pattern characteristic of both species. Thus, some MurchisonRiver, WesternAustralia individuals had distinct spots that had not The Murchison River is located in the Pilbara (or Indian Ocean) Drainage Division of Western coalesced into bands on the snout and cheek (characteristicof L. unicolor)buthad yellowish fins Astuasntdraarldisa,adnrdainisinglaragpeprboyximWaetsetleyrn12A0us0t0r0alkiman2 (characteristicofA. caudavittata) andeitherhadno, or one or two weak or strong band(s) on their a(Fpipgruorxeim3)a.teIltsyh5e0a0dwkamterisnlaarnidsefnreoamrtMheeekmaotuhatrhraa,t cAa.udcaauldalvoibtetsat(ac,haroarctienrtiestrimcesdioafteeitbheetrwLe.enunitchoelotrwoo)r Kalbarri. Precipitation near the river mouth is (see Figure 1). It was suspected that these rmeml/atainvneluym)l,owwiatnhdlohwigshulymmseerasaonndalme(mdeiaunmcwaint37e5r individualswerehybrids. falls, while the inland reaches receive relatively marginal and unpredictable rainfall (ca 238 mm/ annum) throughout the year (data provided by the Western Australian Bureau of Meteorology). The riverismarginallysalinethroughoutitslength. Samplinglocalitiesandenvironmentalvariables Thirteen sites on the lower, middle and upper Murchison River were sampled using a variety of seine nets and rod and line. The seine nets were comprised of either 3 or 6 mm woven mesh. The main study sites (3 and 4) were situated at the easternend oftheKalbarri National Park(Figure3) and the water temperature and conductivity were recorded in each month between December 2000 and November 2001 at site 4. Other sites were sampled on a single occasion and the numbers of each species (that superficially resembled either L. unicolororA. caudavittata)wererecordedandthey supraneurals, anterior proximal pterygiophores, anterior neural spines and were then released immediately. Fish that were anterior Is' dorsal spines in A, keptwereeuthanased inaniceslurry. Leiopotheraponand B, Amniataba. N.B.One supraneural anterior to first neural spine in Morphologyandosteology Leiopotherapon(versus2 in Amniataba)', one Each fish retained from theMurchison River was sdLoeercisooapnlodtsphaiennredapootnnhif(rirvdsetraasnnuttseer2riiooinrrApptmteneriryaygtgiaiobpoahpo)h;roaerneidsn msuebajseucrteivdetoastsheesnsemaernetstm1amdem a(stottaol lwehnegtthh)erandthea insertingbetween successiveneural spines in individual superficially resembled more closely L. Leiopotherapon(fromVari1978). unicoloror A. caudavittata; based on fin and body 136 D.L.Morgan,H.S.Gill coloration and patterning, and also overall shape dorsal-fin pterygiophore; the position of the (in L. unicolor the body is slender and the snout insertion of the first proximal dorsal-fin rounded, whereas the body of A. caudavittata is pterygiophore (between second and third or first moderately deep and the snout pointed and and second neural spine); and number of anterior relatively shorter than that of L. unicolor). The proximal dorsal-fin pterygiophores inserting vertebral column, and associated supraneurals, betweensuccessiveneuralspines (oneortwo) (Vari pterygiophores and dorsal spines, below the 1978, and Figure 2). The vertebral column and anterior-most section of the first dorsal-fin were associated structures werethen drawn with the aid then displayed by dissection and the following of a dissecting microscope and compared to those characteristicsdetermined: numberofsupraneurals that comply with either L. unicolor or A. in front of the first neural spine (one or two for caudavittata (Vari 1978, and Figures 2 and 4). Leiopotherapon and Amniataba, respectively); Dissection, rather than radiography, was used to number of dorsal-fin spines (one or two) on first determine the osteolgy of the first dorsal-fin as a Figure3 The rivers and embayments (Shark Bay) in Western Australia from which terapontids were examined. The mainstudysitesontheMurchisonRivers,andthemajortowns,KalbarriandMeekatharra,arealsogiven. HybridisationofL. unicolorandA. caudavittata 137 preliminary investigation demonstrated that Reproductivebiology radiographs were often difficult to interpret, The gonads of each fish retained from the whereas dissection provided unambiguous and Murchison River were examined, and the sex and accuratedescriptions. stage of gonadal development determined The osteology of individuals of presumptive L. macroscopically.Thestageofgonadaldevelopment unicolorfrom the Gascoyne (5 specimens), Minilya was based on the following criteria adapted from (4), Cane (5), Yule (5), De Grey (5) and Fitzroy (5) Laevastu (1965): stage I/II (immature); stages III/IV rivers, andA. caudavittatafrom theSwanRiver(5), (maturing); stage V (mature); stage VI (spawning) Shark Bay (25) and Cape Keraudien (33) were also andstageVII(spentorrecentlyspawned). determined (Figure 3). Specimens of Leiopotherapon unicolor in the collections of the Western Australian Museum were also examined RESULTS mm (P4336, 4, 56-83 TL, Murchison River, 12 mm Environmentalvariables ORicvteorb,eGral1e9n58a;,P2573S6e2,pt3e,m7b5e-r1019061). TL, Murchison The mean water temperature at the main study A(F.Iingcuaorureddae4vr)ittmtooatrdaeettechrleomsimenileyniwrhmeesutemhmebnrlueamdbveLa.rriuaonnfitccpohalatontrgeerosnr sFAieutbgersuusaftroyl(lc(aocwa1e6d2°7C°a)C)s2e0aa0n1sd.onTarlehaecphacitontngedrunac,tmivpiientaiykmiounfmgthiienn required to transform each individual pattern to main study sites throughout the year of the study that of L. unicolor and A. caudavittata was ranged from 1.6-26.8 mScnr1, but averaged ca 13.2 mScnr'. estimated. In these transformations the following assumptions were made. Firstly, in general, dorsal spines and pterygiophores were not independent, Distributionofterapontidsinthe Murchison thus the loss, gain, or movement of an individual River pterygiophore included a corresponding loss, gain Terapontids were captured at 13 sites in the or movement of the associated spine. Secondly, in Murchison River. Forms that superficially more the case of the first pterygiophore, it was closelyresembled L. unicolorwerefoundatallsites considered that the loss of a second spine on that except the two most downstream locations (Figure pterygiophore could occur without losing the 3). Those resembling more closely A. caudavittata pterygiophore and other spine, and that gaining a were captured at nine sites, ranging from the two second spine on the first pterygiophore did not downstream sites to the most inland sites. Both require gaining an additional pterygiophore 'forms' co-occurred atsevensites. bearing two spines. It is worth noting that if this second assumption is not made, and Morphologyandosteology transformations require the loss (or gain) of the Ofthe231 terapontidsexaminedforvertebraland whole pterygiophore they contain either the same first dorsal osteological comparison during this number of steps or require additional steps. study, 32different patterns were evident(Figures2 Finally, wherethere aretwoor moreequallyshort and4,Table1). transformations, the one presented in Table 1 In the Murchison River samples collected during maximises movements rather than losses and the current study, only 24 (30%) of the 80 gains, e.g., in Type 19 the transformation to individuals that superficially more closely Leiopotherapon presented is: Step 1, move resembled L. unicolor had an osteological supraneural 2 to between neural spines 1 and 2; configuration that complies with Vari's (1978) Step 2, move supraneural 3 to between neural description for Leiopotherapon (see also Figures 2 spines 1 and 2; Step 3, remove 1st pterygiophore; and4). Intheremaining56individuals,20different Step 4, remove 2nd or 3rd pterygiophore; Step 5, osteological patterns unlikethosedescribed forany move new pterygiophore 1 to between neural other terapontid, were found (Figure 4, Table 1). spines 2 and 3; Step 6, move new pterygiophore 2 The most common of these variant configurations to between neural spines 3 and 4. The alternative (Type 4, 20 individuals) differs from the is: Step 1, move supraneural 2 to between neural configuration diagnostic of Leiopotherapon only in spines 1 and 2; Step 2, move supraneural 3 to having the second pterygiophore arising from between neural spines 1 and 2; Step 3, remove 1st between thesecond and third, ratherthan thethird pterygiophore; Step 4, remove 2nd pterygiophore; and fourth, neural spines (Figure4,Table 1). Ofthe Step 5, remove 3rd pterygiophore; Step 6, gain remainingconfigurations present in specimens that additional pterygiophore between nemal spines 3 resembled L. unicolor, types 1, 2, 3, and 5 alsovary and 4. Thus, the first transformation comprises from the pattern diagnostic of Leiopotherapon in four movements and two losses, whereas the only requiring the addition or movement of a alternativecomprisestwomovements, threelosses supraneural or pterygiophore and its spine, whilst and oneaddition. Type 7 only requires the addition of a supraneural 138 D.L.Morgan,H.S.Gill Type 1 -Lu Type2-Lu ywy y Type3-Lu Type 13 -Lu Type23 - Lu w/ v y Type4-Lu Type 14-Lu Type24-Ac w/ r/wi y y Tffl'V/l Type 15 -Lu Type25 -Lu Type29-Lu&Ac w/M Type 10-Lu v/ttf Figure4 Diagramatic representationofthe osteology oftheanterior sectionofthe first dorsal fin for the teraponids examinedandwhethertheysuperficiallyresembled Leiopotheraponunicolor(Lu)orAmniatabacaudavittata (Ac). , ; , Hybridisationof L. unicolorandA. caudavittata 139 and a slight increase in the size of the first description, two were Type 4 (one step) and one supraneural. The remaining patterns (types 8, 10, was Type 11 (two steps). Of the five L. unicolor 12, 13, 15, 17-23, 25 and 29) require between two examined from theFitzroy River one followed Vari and six transformation steps. Of these patterns, (1978), while the others were types 4 (two, one types 1-5, 7,8, 10, 12and 13areclosertothepattern step), 11 (oneindividual, two steps) or 15 (one, two diagnostic of Leiopotherapon, types 18-23, 25 and steps).TheType11 pattern wasalsofound ina fish 29areclosertothatofAmniataba,and types 15and from the Murchison River, but in that case the 17areequidistantbetween thetwo. individual more closely resembled Amniataba Of the 52 individuals that superficially more (Figure 4, Table 1). All of these patterns are equal closely resembled A. caudavittata, 42 (ca 81%) had (Type 15) or closer to that diagnostic of theosteological patterncharacteristicforthatgenus LeiopotheraponratherthanthatofAmniataba. (Figure 2) (Vari 1978), however, the remaining 10 All A. caudavittata examined from the Swan (ca 19%) displayed nine different osteological River(5) andSharkBay(25), and30of33examined configurations unlike those described for any other from Cape Keraudien had a dorsal osteology that terapontid (types 6, 11, 16, 24 and 26-30 in Figure conformed to Vari's (1978) description for 4). Three patterns (28-30) are only one Amniataba. The three individuals that differed in transformation step from that diagnostic of osteological formula from Vari's (1978) description Amniataba whilst the remainder are either two were either types 28 (1) or 30 (2). Both of these (types 16, 24, 26 and 27) or three (types 6 and 11) patterns only require the movement of a single steps away. Of these patterns, types 24 and 26-30 supraneural to conform to the Vari's (1978) are closer to the pattern diagnostic of Amniataba, diagnosisforAmniataba(Figures1 and4,Table1). types 6and 11 arecloser to thatof Leiopotherapon, Based on comparisons of the minimum number andType16isequidistantbetweenthetwo. of steps required for a pattern to transform into Of the L. unicolorexamined from the collections thosediagnosticofLeiopotheraponandAmniataba, of the Western Australian Museum, and that all thefishthatsuperficiallymorecloselyresembled L. superficially resembled thatspecies,twospecimens unicolorbore 13 patterns thatwere more similar to from 1958 (P4336) conformed to the diagnostic the diagnostic configuration for Leiopotherapon pattern ofthegenus andtheremainingtwo wereof (i.e., types 1-5 and 7-14), eight that were more theType2pattern, i.e.,onestepfrom thediagnostic similar to that diagnostic for Amniataba (i.e., types pattern, whilst two specimens from 1961 (P5362) 18-23, 25 and 29), whilst two are equidistant were Type 4 and the other was Type 10, i.e., one between the two diagnostic patterns (i.e., types 15 and two steps from the diagnostic pattern, and 17)(Figures2and4,Table1). Inthecaseoffish respectively (Figures 1 and 2, Table 1). All of these resemblingA. caudavittata therewerefivepatterns patterns are closer to that diagnostic of that more closely resembled that diagnostic for LeiopotheraponratherthanthatofAmniataba. Amniataba (i.e., types 24, 26-28 and 30), two were Of the five L. unicolor examined from the moresimilar to thatof Leiopotherapon (i.e., types6 Gascoyne River, one conformed to Vari's (1978) and 11) and the remaining pattern was equidistant description,whilsttheothersweretypes4 (2)and2 between the two (i.e., Type 16) (Figures 2 and 4, (2), patterns that are both only one transformation Table1). step away from the diagnostic pattern. All four of Vari (1978) noted that Leiopotherapon and thepresumptive L. unicolorfrom the Minilya River Amniataba can be distinguished by the following had a dorsal osteological pattern unlike that osteological criteria: one supraneural bone in front described by Vari (1978), oneof thesehad a Type2 of the first neural spine in Leiopotherapon versus pattern (one step from the diagnostic pattern) that two in Amniataba one spine on the first dorsal was also found in fish from the Murchison and pterygiophore versus two; the insertion ofthe first Gascoyne Rivers, one had a Type 9 pattern (two proximal dorsal pterygiophore between the second steps) and the remaining two individuals had a and thirdneuralspinesversusitsinsertionbetween Type 14 configuration (three steps), these latter the first and second; two of the anterior proximal patterns were unlike any of the other individuals dorsal pterygiophores insertingbetween successive examined (Figure 4, Table 1). One of the five neural spines versus one. In the current study, of individuals from the Cane River conformed to L. the individuals that superficially most closely unicolor, whilsttheremainingfourall exhibitedthe resembled L, unicolor, six had two supraneurals in Tyne 4 pattern (one step). In the specimens from front of the first neural spine (types 15 (3), 20, 23 the Yule River, three individuals had the pattern and29),fivehadtwospines(types 14(2), 18,25and diagnosticof L. unicolor, onehad aType4and one 29), seven had the first proximal dorsal a Type 12 pattern, i.e. one and three steps away pterygiophore inserting between the first and from the diagnostic pattern, respectively (Figures 2 second neural spines (types 20, 21 (4), 22 and 23) and 4, Table 1). Of the five L. unicolorexamined in and six individuals had only one of the anterior the De Grey River, two conformed to Vari's (1978) proximal dorsal pterygiophores inserting between , , ^ T 140 D.L. Morgan,H.S.Gill -a g I! 3 o ^ > tj ^Omtc0ge,£2rCae —2xrr..>—2>3- —•2Eno «^5 (N (N S0Oi-cc5dur o«C uIacrceS—z>cG, 2G £ CL C2Q/G)1 c^2c^-n5, C20G/11 ^ "5 -psrPecXx"LSGc3 -*~wsoC3»rX52—e5.X^UX.C«gciciTI*-x-S*C&u3r£2eL '2%oO3 #GC2CD tTC<2r/:2Ne1 ((H2CC/NNL1 C2THC•Q/M—ll7l1 2HCCC/NLl 52T(Cc/rZ^Nl5e5 C ^ Xccu >,S .2 03 GOJ '>0C3«C/3 §-S In TT^g0GO0331 c7^a•/o5c£I“I;5•>vjX"=<Q°Q6J-oo/rJ3i,uXoOcf3 £1^-£Gocl3m- c2•coGnJ "1^hP33-l CJC2TC2l/O/aCOll»1v-rCr2r0;eglbm. C2VO0/hl C2PHC/h^l" C2r5C/r—Lle c2HHC/Ll ""C'2rc/OOel CC2100c/N11l C2CCr2/—OOl 2CC2OO CL• *c2-co2 CH2COL C2r«u—O C5-/i c2rsn- •£toCPol0Sn.1 1xH.t§1«21KSL^--vaatO«i>»o |-U4G2°cIsIiIuiX|%^c Groe ^p^< t2DcrNe u-co2,5i —^as (-2N8 Z 2 2 g c 2irOcne —o[2-j£_ r2C—PrO'/l3a «n—-—tC,2o o 5 -±cO5*-2* c^Gi03=3lbao£"i 'oursq-j5i^X[i||-Jf^ori:e in a ^ Xi ro ^1 -ag o1 -«a 52- C./,3 .,2n c CJOS !? -CO o0Xi1 o Z ^ to00Cn1 T*S*j3. o.cOuJ 5^r9-e KCS>Q3 ^"'c3ii gfj* § 05 Qi O r~ G 2to ._2q S3 c3n: faS Po §• IT ^ ^2 CO CO/3 •Q£J 2^2 C2D £?£3G£3 ^2‘£cr5n45 •.^--S0GCuO**f£O2«£t-G23J-eO2-H 4Xt&£"2.<Piu^f3^£-:~33l^n,)£UDf1x“C22r>rN^o:rt1h1]“',lX^-X,""«C£C°^1£«pc0^J*P//£~)233Ll O Xc2C<00G00iONo1111 XCC2C00c00ONO1111 XC"XX2Cr0r§ON1e1e HCoXCGO0•£NLc11— 'HXC’0Nu5oO115 "c;’Xc2rcrGr’o5o0leeeS XcCHXCc2^0•NoNoT1 XC^C22C0c00OOO111J CCX2HX|G00OO11 .E£ --22-"6 ^2 °^•§ N'U£3 > £H jO «m 5| UD ’35 S'O?'rl s s o' (j 60 EirlD 1OS 2C£L Ho>>t><> HG>., H^g-55< C^>HNV HCL HCL HCl CL< hC"L 5 1 ' HybridisationofL. unicolorandA. caudavittata 141 * £ H £QJ HCL £ZJ <£> -T33 £QJ CN *T53 CL CD CN CN CzD CfNi3 CN EQJ CN &2. c 1 CzD rH T— rH QJ rH tH— r- o zCD <-. CzD CzD >5 CzD CzD CL CZD NS S£ oc >5 between 3to Z 2 SN 1- 2; CN Dl CN <N & 2w C2D 2w CzD 1 rZn a?> C2D a) ZIidS •C*D zc NS xO> *H-* ^O rHC-l CD CzD .C•COZM2DDJ CCZ»DDa X^j^£°i>> C2cCIDD •ZZO“JJ HCff-Ltt1 b3tetwoeen N34&S. S Z CN ^CLL- ^ H^ co3o 1rH4 <HCNL £ C CZO -H0Ca-)1 "'* fc\n| Z««a tzn 1&2SN;between 7ZB£~eHCHam. Zieo ^27(2s$aN'! ‘^-c£a'SBi,Xf<Ct|*oi_ni S* *s ICZ21£§5D1 £phcOs-- yi_I2<z£>-_e^>D?4)^*C^hTf£Q£s-EJz|n-^0OIa5 r™ZCIoIDIHCCUCmD 2C25Q£DJI -3-g*QCDI-Jl* T2c.-D5d*I: °C-(H*2|N8lQ' cmZ2cCz£nDJj -Hn,H^C2El-o-h!iZL^*Cx2r3Dl- HCCCD0EcD)nDl CZCc+£CDDd-ID C(HCCQQ£DNlJJ bS2N4ettwSNeo-en a1PPddirtiTeTo;mna.l 2, ? s M; (2, 30 CL CL CL CL CL D. Cl CL CL h" H H>-, H' H H' H U H>, TypeCK) 142 D.L.Morgan,H.S.Gill successive neural spines (types 5 (2), 9, 14 (2) and specieswithinthesetwogenera, Variradiographed 17). Inthecaseofindividuals thatsuperficiallymost or cleared and stained for osteological examination A closelyresembled caudavittata,fivehad onlyone (Vari personal communication), given such a high supraneural in front ofthe firstneural spine (types occurrenceofdifferent pterygiophore patterns (i.e., 6, 11, 16 and 28 (2)), four had the first proximal 54% were unusual), it could be expected that he dorsal pterygiophore inserting between the second would havenoticed someexceptions.However, our and third neural spines (types 6, 11, 16 and24) and preliminary use of radiographs suggested that it four individuals had two of the anterior proximal was often difficult to discern minor variations in dorsal pterygiophores insertingbetween successive pattern usingthismethod.Thus,althoughnotnoted neural spines (types 6, 16, 26 and 27). All by Vari (1978), it is likely that some minor natural individuals that superficially resembled A. variations in pterygiophore pattern occur in these caudavittata had two spines on the first dorsal species. For example, of the 63 specimens of A. pterygiophore. caudavittataexamined from theSwan River, Shark Bayand Cape Keraudien, i.e., estuarineand marine Reproductivebiology sites at which L. unicolordoes not exist, the three Individuals exhibiting characteristics of both that did not conform to the pattern diagnostic for species had lengths ranging from 39-253 mm TL, that genus only differed in the position of the while those representing L. unicolor and A. secondorthirdsupraneural.Furthermore, although caudavittata ranged in length from 47-211 and 29- variations in pterygiophorepatternwere evident in 206 mm, respectively. Of the 13 fish greater than some specimens of L. unicolorexamined from the 200 mm TL, 11 exhibited characteristics of both Western Australian Museum, these variations were species. minorand all specimens exhibited the coloration of Fromthemacroscopicstagingofthegonadsofall thatspecies. fish dissected, including those with characteristics Notwithstanding that some minor differences in of both species, it was evident that many were the anterior dorsal-fin pterygiophore pattern occur either mature, spawning or had recently spawned. naturally, the level of variation in the first dorsal/ The peak spawning period for the two species and neural spine patterns and the fact that many the fish exhibiting characteristics of both species in specimens exhibited coloration and shape theMurchisonRiverappeared tobein latesummer/ characteristics of both L. unicolor and A. earlyautumn. caudavittata suggests that in the Murchison River The sex ratio in the individuals that exhibited these two species can readily hybridise. This is both characteristics was ca 1.8 females:l male, further highlighted by the fact that in many whereas for L. unicolorand A. caudavittata it was instancesitwasverydifficulttodecidewhetherfish 2:1 and 1.6:1, respectively. superficially resembled L. unicolor or A. caudavittata. Furthermore, as only seven of the 25 presumptive L. unicolor examined from the DISCUSSION Minilya, Cane, Yule, De Grey and Fitzroy Rivers Ofthe 161 terapontids collected from the riverine were actually L. unicolor as described by Vari sites during thecurrent study, thecharacteristicsof (1978), andaseach riverhad at least twospecimens the first dorsal/neural spine matrix resembled L. that did not conform to the diagnostic pattern, it is unicolorin about 20% of cases and A. caudavittata possible that hybridisation between terapontids is in about 26% of cases. The remaining 54% of fish not restricted to the Murchison River. It is also had first dorsal/neural spine osteological worth noting that the related barred grunter characteristics unlike any of those described for (Amniataba percoides) is present in some of these terapontids by Vari (1978). Whilst many of the rivers and that Dr Barry Hutchins of the Western novel patterns described in the current study are Australian Museum believes that hybridisation very similar to those described by Vari (1978) as between terapontids may be quite common in the diagnostic for the two genera, others bear little rivers of the Kimberley (Hutchins pers. com., see resemblance to the diagnostic patterns. Vari (1978), alsoHutchins 1981). who examined 10 specimens from the Murchison In addition to proposing that much of the River,76fromtheDeGreyRiverand over 100from variation in coloration and anterior pterygiophore the Fitzroy River, made no mention of the patterns described in the current study is the result osteological variability evident in thecurrentstudy of hybridisation, we also suggest that, in the for these species. Furthermore, he noted no Murchison Riveratleast, (1)hybridisationhas been variation in the patterns of other members ofthese occurring for at least the last few breeding seasons, genera considering the two patterns to be (2) hybrids are reproductively viable and (3) some diagnostic for Leiopotherapon and Amniataba. individuals may have been hybrids of hybrids. Although it is not known how many individual L. Thesefurtherhypothesesarebasedonthefactsthat unicolor or A. caudavittata, or specimens of other fish from the Murchison River that exhibited such