sign in sign up
<<
Home , Alisterus, Bluebonnet (bird), Budgerigar, Cacatuinae, Cyanoramphus, Cyclopsitta

TheCondor 93:302-317 0 Thecooper ornitholcgical society 1991

RELATIONSHIPS AMONG THE AUSTRALO-PAPUAN , LORIKEETS, AND (AVES: PSITTACIFORMES): PROTEIN EVIDENCE ’

L. CHRISTIDIS Department of Ornithology,Division of Natural History, Museum of , Melbourne, Victoria, 3000

R. SCHODDE Australian National Wildlif Collection,Division of Wildltfe and Ecology, CSIRO, Gunghalin, Canberra, Australia 2602

D. D. SHAWAND S. F. MAYNJB Molecular and Population GeneticsGroup, ResearchSchool of Biological Sciences, Australian National University,Canberra, ACT, Australia 2601

Abstract. Allozyme variation at 21 presumptive protein loci was examined by electro- phoresis and used to assessrelationships among Australo-Papuan parrots, lorikeets and cockatoos.Hypotheses of relationships were generated from the data by phenetic and cla- distic analyses.The results,when taken into account with other biochemical, chromosomal and morphological data, demonstrate that cockatoosform a monophyletic lineage distant from the other Australo-Papuan parrots and lorikeets. The lorikeets are also monophyletic, but are clusteredamong other parrots.A core ofAustralian broad-tailed (platycercine)parrots is defined by the rosellasand ringnecks(Platycercus, Barnardius), (Northiella), Red-capped (Purpureicephalus), (Lathamus) and grassparrots (Psepho- tus). New Guinean Psittacellaalso appearsto be a member of this assemblage,to which the polytelitine parrots (-)may be linked as well. Other “conventional” platy- cercine parrots- the Ground Parrot (), (Melopsittacus), Red-fronted (), and Blue-winged and Bourke’s Parrots ()-are still more distant and of disparate affinity; the two latter speciesare polyphyletic among the platycercines.Of psittacineparrots, Pclectus () and Red-cheeked Parrots (Geofioyyus) are closely related but their links to other psittacine genera are not clear. Similarly, the relationships of the fig-parrots (),pygmy-parrots (Micropsitta), (Aga- pornis) and ring-necked () are ambiguous.Biogeographical implications of these results are canvassedin the discussion. Key words: Parrots:protein electrophoresis;; systematics; Australo-Papuan.

INTRODUCTION brush-tongued lorikeets, both of which are con- The order Psittaciformes comprises some 330- fined to the Australasian and South Pacific 350 of parrots, lorikeets and cockatoos regions. The arine parrots, comprising the entire which occur naturally in Central and South complement, also appear to form a America, Australasia and the South Pacific, Af- monophyletic radiation, judged by their wide rica and southern Asia. There are two major ra- biochemical distance from other parrots (Mai- diations, one in Australasia and the other in South nardi 1962, Gysels 1964) and their severalunique America. Although the order is well-defined pigmentary, ontogenetic and copulatory traits morphologically, the primary evolutionary lin- (Smith 1975). eageswithin it are not so clear-cut (cf. Glenny Most controversy centers on the only other 1957, Sibley 1960, Brereton 1963, Boetticher large group, the Australasian -eating psitta- 1964, Sibley and Ahlquist 1972, Smith 1975, tine parrots. Up to eight tribes and subfamilies Homberger 1980). Two discreteassemblages have have been distinguished among them (Smith been recognized in all studies: the cockatoosand 1975, Homberger 1980) and there is consider- able dispute over their composition and rela- tionshipsto the cockatoosand lorikeets(see Smith I Received28 June 1990. Final acceptance6 No- 1975 and Table 1 this study). Towards resolving vember 1990. some of thesequestions, we employed multilocus

[3021 BIOCHEMICAL SYSTEMATICS OF PARROTS 303 protein electrophoresisto examine relationships TABLE 1. Recent classifications of the genera ex- among 36 speciesof Psittaciformes, covering six amined in the present study. of the tribes recognized by Smith (1975). The survey is limited to the Australasian region, and Homberger1980 smith 1975 Peters1937 focussedon the principal loriine and psittacine Cacatuinaea Kakatoeinae assemblages. Cacatuini’ Platycercinae Psittacinae MATERIALS AND METHODS Platycercini Platycercini Protein electrophoresiswas performed on 80 in- Melopsittacus Melopsittacus Melopsittacus dividuals of 36 species(Table 2) representing 16 Neophema Neophema Neophema Psephotus Psephotus Psephotus of the 22 genera of Australo-Papuan psittacine Northiella Northiella Northiella parrots, seven of the 10 genera of lorikeets, and Purpurei- Purpurei- Purpureiceph- two of the five generaof cockatoos(Condon 1975, cephalus cephalus alus Beehler and Finch 1985), as well as one Pacific Platycercus Platycercus Platycercus species (Norfolk Island Red-fronted Parakeet, Barnardius Barnadius Barnard&s Lathamus Lathamus Cyanoramphus) and two Afro-Asian psittacine Cyanoramphus Cyanoramphus Cyanoramphus genera (Psittacula, Agapornis). Locality data for Pezoporus Pezoporus the material collected are available from the au- Psittacinae Psittacinae thors on request. Psittaculini Electrophoresis was carried out on liver and Geojiioyus Geofroyus breast muscle samples which had been stored in Eclectus Eclectus Eclectus Psittacella Psittacella liquid nitrogen. Separatehomogenates of the two Alisterus Alisterus Alisterus tissueswere prepared by grinding a cubic milli- Polytelis Polytelis Polytelis meter of each in 300 ~1 buffer (0.1 M Tris, 1.O Agapornis Agapornis Agapornis mM EDTA, 0.5 pi/ml 2-mercaptoethanol, 0.05 Psittacula Psittacula Psittacula mM NADP, pH 7.0). The homogenates were Loriinaeb Loriinae Loriinaeb Loriinib then spun in an Eppendorf centrifuge for 3 min incl. Lathamus and the supematant screenedfor 25 enzyme sys- Cyclopsitta tems representing 32 presumptive loci (Table 3). Not examined Loriinae not recognized Enzymes were stained according to the recipes Psittaculi- in Harris and Hopkinson (1976) except GOT rostrini (Table 3), for which the procedure of Shaw and Cyclopsitta Prasad(1970) was followed. All systemswere run Not examined Loriinae Micropsittinae in a cellulose acetate matrix on a paper support Micropsittini (Cellogel, Chemetron, Italy). Where two loci rep- Micropsitta Micropsitta resented a single enzyme, the most anodal was a Includesthe 2 8em-a listedunder Cacatuidae in Table 2. bIncludes the 7 generalisted under Lmiidae in Table 2. designated - 1, and the other - 2. Individual al- leles were given alphabetical designations in se- quence from the anode, beginning with “a.” distance-Wagner (Farris 1972, Swofford 198 1) Of the 32 loci screened,the following 10 were dendrograms were then constructed with the excluded from analysis because they could not BIOSYS- 1 program (Swofford and Selander be scored consistently across all species: GPT, 1981). The distance-Wagner dendrogram was GLUD, TPI, ACON- 1, ACON-2, EST- 1, MDH- rooted by both mid-point and out-group pro- 2, GDA, NP, and PGM-2. Variation at LDH-1 cedures, the cockatoos being used as the out- and LDH-2 could not be distinguished unam- group for psittacine and loriine lineages because biguously because of differential expression of of the morphological (Smith 1975, Homberger the polymer bands. Accordingly, their variation 1980) biochemical (Adams et al. 1984, Ovenden was scored on pattern alone and treated as a et al. 1987) and chromosomal (Christidis et al., single locus. in press)evidence that they are a distinctive sister From allelic frequencies at the 2 1 loci remain- lineage of the other Psittaciformes. ing (Table 2) Rogers’ (1972) and Nei’s (1978) A cladistic analysis was also performed by genetic distances were calculated between taxa treating the loci as charactersand their constit- (Table 4). UPGMA (Sneath and Sokal1972) and uent alleles as character states. Where loci were

BIOCHEMICAL SYSTEMATICS OF PARROTS 307

TABLE 3. Enzymes examined, buffers used, and tissue distribution of each enzyme.

Enzyme RUMbIg Running (E.C. No.) Abbreviation No. of loci Tissue buffer time (hr)b

Aconitase ACON 2 Liver, F 3 (4.2.1.3) muscle Adenylate kinase AK 1 Muscle A 3 (2.7.4.3) Aldolase ALD 1 Muscle D 3 (4.1.2.13) Creatine kinase CK 2 Muscle A 3 (2.7.3.2) Enolase ENOL 1 Liver A 3 (4.2.1.11) Esterasec EST 2 Muscle A 1.5 (3.1.1.1) Fumerase FUM 1 Liver F 2.5 (4.2.1.2) General protei& GP 1 Muscle A 3 Glucose-phosphateisomerase GPI 1 Liver E 3 (5.3.1.9) Glutamate dehydrogenase GLUD 1 Muscle A 3 (1.4.1.3) Glutamate oxaloacetatetransaminase GOT 2 Liver F 3 (2.6.1.1) Glutamate pyruvate transaminase GPT 1 Liver F 2.5 (2.6.1.2) Glyceraldehyde-3-phosphatedehydrogenase GA3PD 1 Liver D 3 (1.2.1.12) Glycerophosphatedehydrogenase GPD 1 Liver F 3 (1.1.1.8) Guanine deaminase GDA 1 Liver C 1 (3.5.4.3) Isocitrate dehydrogenase IDH 2 Liver F 3 (1.1.1.42) Lactate dehydrogenase LDH 2 Muscle A, D 3 (1.1.1.27) Malate dehydrogenase MDH 2 Muscle A 1.5 (1.1.37) Mannose phosphate isomerase MPl 1 Muscle C 1.5 (5.3.1.8) Phosphoglucomutase PGM 2 Liver A 3 (2.7.5.1) 6-Phosphogluconatedehydrogenase 6PGD 1 Liver C 2.5 (1.1.1.44) Phosphoglyceratekinase PGK 1 Liver A 3 (2.7.23) Purine nucleosidephosphorylase NP 1 Liver F 1.5 (2.4.2.1) Pyruvate kinase PK 1 Muscle B 2 (2.7.1.40) Triose-phoshpate isomerase TPI 1 Liver D 3 (5.3.1.1)

a A = 50 mM TEM, B = 15 mM TEB,,C = 50 mM TEM + NADP, D = 50 mM TEM + NAD, E = 25 mM TEB, F = 0.1 M Tris-citrate. Recipes for 1 liter of above buffers. A: 6.06 g Tns, 1.86 g Na EDTA, 0.20 g anhydrous M pH to 7.8 with Maleic acid. B: 1.82 g Ttis, 1.86 g Na EDTA, 0.20 g anbydrous M&l, pH to 8.0 with boric acid. C: as for A but add 10 mg N p DP. D: as for A but add 10 mg NAD. E: 3.06 g Tris, 1.86 g Na EDTA, 0.20 g anhydrous MgCI, pH to 8.0 with boric acid. F: 12.11 g Tris, pH to 7.8 with citric acid. b At 7 mA per 12 cm gel (except B and E buffers; 5 mA). =By method A in Han-is & Hopkinson (1976) wth 4-methyl-umbelliferyl-acetate. d Stained with amide black.

310 L. CHRISTIDIS ET AL. polymorphic within a species,the allele in higher Nei D of only 0.04. At distances ranging from frequency was designatedas the state for the tax- 0.26 to 0.45, the New Guinean tiger-parrots on. Alternate alleles were equal in frequency in (Psittacella) were also close to this assemblage. seven instances. Where one or both of the alleles Other slight intergenetic distanceswere recorded were autapomorphic, an autapomorph was ex- between the polytelitine king parrots, Polytelis cluded because this would not affect cladistic and Alisterus,at 0.05, and between the “typical” analysis; where they were not, we made a con- palaeotropic parrots, Eclectusand Geofloyus, at servative decision to choose as the state for the 0.29. Conversely, the two speciesof platycercine taxa concerned that allele which was most fre- grass parrot, Neophema bourkii and N. chryso- quent and widespread in other parrots. stoma, differed by 0.55. The data for cladistic analysis were then en- UPGMA phenograms based on Nei’s (1978) tered as unordered, without weighting, into the and Rogers’ (1972) measures were identical but PAUP program (Swofford 1985) which con- with one exception. Nei’s D clustered the Blue- structs phylogenies according to the principle of winged Parrot (N. chrysostoma)with the fig-par- maximum parsimony. Mid-point and out-group rots (Cyclopsitta), pygmy-parrots (Micropsitta) procedures using the cockatoos were employed and lorikeets, while Rogers’ D grouped it with again to root the trees. To reduce arbitrary res- the Red-fronted Parakeet (Cyanoramphus) near olution of polychotomies as distinct trees, taxa the Australian broad-tailed assemblage (Platy- which sharedidentical allelic constitutions or dif- cercusto Psephotus).The Rogers based pheno- fered only by autapomorphies were excluded gram had a higher copheneticcorrelation of 0.92 1 (Swofford 1985). Those excluded from this anal- compared with 0.893 for the Nei phenogram. ysis are asterisked in Table 2. Also excluded, to Accordingly, the phenogram based on Rogers’ improve the efficiency of the program run, were (1972) D is depicted in Figure 1. In it, the major the following loci in which variation was limited dichotomy is between the cockatoos and other to a single or species: GA3PD, FUM-1, parrots. The two main cacatuine lineages, the MDH-1 and ALD. Because many (50) equally black cockatoos(Calyptorhynchus) and white and parsimonious trees were produced, consensus pink cockatoos () are also well differ- trees were constructed with the “strict” method entiated. The first taxon split off among the par- (Rohlf 1982). rots themselves is Bourke’s Parrot, Neophema bourkii. The next dichotomy divides the re- RESULTS maining parrots into two major assemblages.One Phenetic analysis of geneticdistance data. Of the includes all lorikeets, and the fig- and pygmy- 2 1 loci assessed,none were monomorphic across parrots (Cyclopsitta, Micropsitta) as sister lin- the range of speciesexamined (Table 2). Rogers’ eages.The Australian Budgerigar(Melopsittacus) (1972) and Nei’s (1978) genetic distances were and African lovebirds (Agapornis)are also linked correlatedthroughout (Table 4); accordingly,only distantly to this group. The other major assem- the latter are evaluated below as they are the blage comprises two principal clusters. One in- most commonly used (Avise and Aquadro 1982). cludes the “typical” palaeotropic parrots: Geof- Genetic distances were greatest between the froyus (Red-cheeked and Blue-collared Parrots), cockatoos and the remaining parrots, at values Eclectus and Psittacula (Asian ring-necked par- ranging between 0.80 to 1.65. Within cockatoos akeets). The second comprises the polytelitine and parrots, distances among morphologically king parrots (Polytelis,Alisterus) and core mem- well-separated genera ranged around 0.40 or bers of the Australian platycercines (Crimson higher, similar to the distance between lorikeets , Platycercus,to Red-rumped Parrot, Pse- and other parrots. There were, nevertheless, ex- photus). Linked to them are the Red-fronted Par- ceptions. The seven genera of lorikeets were akeet (Cyanoramphus), Blue-winged Parrot themselvesvery close,being separatedby genetic (Neophema chrysostoma), Ground Parrot (Pe- distancesof only 0 to 0.06. Moreover, other gen- zoporus) and New Guinean tiger-parrots (Psit- era of Australian broad-tailed parrots (Platycer- tacella). cinae) were separatedby distancesof 0.20 or less: Phylogeneticanalysis of genetic distancedata. Platycercus, Barnardius, Northiella, Lathamus, Both midpoint and outgroup (Fig. 2) rooted dis- Purpureicephalus, and Psephotus. Of these, tance-Wagner trees had identical topologies and Platycercusand Barnardius were closest with a lengths (4.668) with cophenetic correlations of BIOCHEMICAL SYSTEMATICS OF PARROTS 311

Aiisterus scapularis Polytelis swainsonii Cymoramphus novaezelandiae Neophema chrysastoma Platycercus elegans r Platycercus eximius Bamardius zonarius Northiella haematogaster Purpureicephalus spurius Lathamus discolor Psephotus haematonotus Pezoporus wallicus

II ’ I I I I I I I I t 1 I I I I I 0.72 0.64 0.66 0.46 0.40 0.32 0.24 0.16 0.06 0.00 Rogers ’ (1972) Genetic distance FIGURE 1. UPGMA phenogramfor representativeAustralo-Papuan Psittaciformes based on Rogers(1972)’ distancemeasure.

0.906. The outlying lineages among the parrots tailed parrots (Platycercus, Barnard&, Purpu- are the polytelitine king parrots and Cyanorum- reicephalus,Northiella and Lathamus), and (3) phus. At the next dichotomy, the core members the Red-rumped Parrot (Psephotus). Further- of the Australian broad-tailed parrots, including more, 46 of the initial 50 trees linked the New the New Guinean tiger-parrots, are clustered on Guinean tiger-parrots (Psittacella) and Red- one side away from the remaining parrots and fronted Parrakeet (Cyanoramphus)with the Alis- lorikeets. Geoffroyusand Eclectusare again clus- terus-Platycercus-Psephotusclade (Fig. 3b). tered togetherbut distant from Psittaculu(cf. Fig. DISCUSSION 1). Phylogeneticanalysis of allelic states. PAUP Concordanceof lineages among the algorithms. analysis with either midpoint or out-group root- The many discrepanciesin the positions of taxa ing produced 50 equally parsimonious trees each among UPGMA, distance-Wagner and PAUP with a length of 95 stepsand a consistencyindex analysesare due to differing assumptionsimplicit of 0.821. A “strict” consensus tree calculated in these algorithms. In the UPGMA computa- from them is depicted in Figure 3a. Two major tion (Fig. l), which assumes a constant rate of clades are evident. The first comprises the fig- evolution among lineages,Bourke ’s Parrot is sep- parrotsand the palaeotropicparrots, Eclectus and arated as a sister group to the remaining parrots. Geofloyus, the latter two as sister taxa. The sec- Where differing rates of protein changeare taken ond clade is subdivided into three lineages: (1) into account, as in the distance-Wagner proce- the polytelitine king parrots (Alisterus-Polytelis), dure (Fig. 2), Bourke’s Parrot clusterswith other (2) most core members of the Australian broad- parrots. Moreover, the comparatively low num- 312 L. CHRISTIDIS ET AL,.

Calyptorhynchusfunereus Cacatua roseicapilla Cacatua gaferita Jrichoglassushaematodus Glossopsittaconcinna Psftteutelesversicolor musschenbroekii Neopsittacuspuflicauda

Oreopsittacusarfakf

I-Pezoporus wallicus Melopsittacus undulatus Micropsittabruijrtii L-P- Micropsittapusio I Cyc/opsittagulielmfterti Psittacula himalayana

Notihiella haematogaster

Lathamus discolor

Psittacellabrehmii -IL Cyanoramphusnovaezelandiae Alisterusscapularis Po/yfe/isswainsonii FIGURE 2. DistanceWagner tree for representativeAustralo-Papuan Psittaciformes based on Rogers(1972)’ distancemeasure, rooted by the cacatuines(cockatoos).

ber of resolved loci may exaggerateinter-locus ringnecks (Platycercus, Barnardius), Bluebonnet variance in estimated genetic distances, thereby (Northiella), Red-capped Parrot (Purpureicepha- magnifying apparent heterogeneity in rates of lus), Swift Parrot (Lathamus) and grassparrots change. In a similar study on passerines,Chris- (Psephotus). New Guinean Psittacella is also a tidis and Schodde (in press) suggestedthat con- member of this assemblage. fidence could be placed only on assemblagesthat (4) Other conventional members of the platy- are grouped consistently by both genetic distance cercines-the Ground Parrot (Pezoporus), Bud- and allelic state data. Accordingly, the following gerigar (Melopsittacus), Red-fronted Parakeet conclusions may be drawn from the data. (Cyanoramphus), Blue-winged and Bourke’s (1) The cockatoosform a monophyletic lineage Parrots (Neophema)-are much more distant and distant from all other Australo-Papuan psitta- of disparate affinity in their allozymes. tine parrots. (5) The polytelitine parrots (Alisterus-Polyte- (2) The lorikeets also form a monophyletic lin- lis) are a sister group of the platycercine parrots. eage but one that is much more close-knit than (6) The Eclectus (Eclectus) and Red-cheeked the cockatoosand which may fall within the Aus- and Blue-collared Parrots (Geofloyus) are the tralo-Papuan psittacid assemblage. only palaeotropic genera to cluster consistently. (3) A core of Australian broad-tailed (platy- The fig-parrots (Cyclopsitta) may also be linked cercine) parrots is defined by the and to them. BIOCHEMICAL SYSTEMATICS OF PARROTS 313

I Psephotus PaZO/XWS Melopsittacus Cyanoramphus Neophema bwrkii Neophema chrysostoma Gecdfroyus I Eclectus Cyclopsitta Psittacella Micropsitta Psitlatula himalayana I a Psinacula derbiana Agapornis

FIGURE 3. a) Strict consensustree for representativeAustralo-Papuan Psittaciformes derived from PAUP analysisof the electromorphsusing cacatuines as the outgroup.b) relationshipswithin playtycercineassemblage of 3a when 4 outlyingPAUP treesare excludedfrom the consensustree.

(7) The pygmy-parrots (Micropsitta) and the are ancient rather than rapidly evolving in their Afro-Asian generaPsittacula and Agapornishave alleles. Karyotypes further stressthe schism be- no obvious links with other taxa screenedhere. tween the cockatoos (represented by Cacatua, Concordanceof lineages with other biochemi- Leptolophus= Nymphicusand Calyptorhynchus) cal, chromosomaland morphologicaldata. Com- and other parrots. Whereas nearly all parrots and parison of these results with other recent protein the the lorikeets have a diploid complement of (Adams et al. 1984, Ovenden et al. 1987) and 60-72 chromosomes, including usually five to chromosomal (Van Dongen and De Boer 1984, seven pairs ofbi-armed macrochromosomes,the Schmutz and Prus 1987, Christidis et al., in press) cockatooshave a higher diploid number (7 2-80) studies, and with the contemporary morpholog- composed largely of single-armed elements ical reviews of Smith (1975) and Homberger (Christidis et al., in press).When taken into ac- (1980), shedsfurther light on relationshipsamong count with the many morphological and behav- the Australo-Papuan Psittaciformes (cf. Table 1). ioral differencesbetween the cockatoosand other Adams et al. (1984) also concluded that the parrots (Smith 1975, Homberger 1980, Adams cockatoosare a distinct lineage among the order, et al. 1984) these data lend support for recog- distant from all other groups in the Australo- nizing the cockatoosas a family. Papuan region. There are no links with the platy- The lorikeets also cluster as a monophyletic cercine broadtails (cf. Smith 1975). Genetic dis- group on morphological and behavioral char- tances between major lineages of the cockatoos acters (Smith 1975, Homberger 1980) and are are as great as between tribal groupings in other linked with the psittacine parrot assemblageon parrots, suggesting,by implication in the dis- karyotype morphology (Christidis et al., in press). tance-Wagner analysis (Fig. 2), that the lineages Although none of these studies clarify their clos- 314 L. CHRISTIDIS ET AL.

Calyptorhynchus / Leptolophus I Glossopsitta Purpuraicephalus Barnardius Psephotushaematonotus Northiella Psephotusvarius Cyanoramphus

FIGURE 4. Strict consensustree for AustralianPsittaciformes. Computed from the data of Ovenden(1984), usingPAUP with mid-point rooting.

est relatives, the lorikeets were perceived by Im- rot (Pezoporus), Budgerigar (Melopsittacus) and melmann (1966) and Holyoak (1973) as spring- Bourke’s and Blue-winged Parrots (Neophema) ing from the Australo-Papuan platycercine cf. Table 1. If these genera are broadtails, then parrots. Indeed, from a single UPGMA com- they are widely divergent elements, as is evident putation, Ovenden et al. (1987) reported that also in their morphology. Unlike other platycer- Plutycercus itselfwas not a member of the “broad- tines which have a type A-2 carotid formula, tail” assemblage but aligned with the lorikeet Melopsittacus and Neophema possessthe ances- Glossopsitta instead. Such a relationship is so at tral type A-l (Glenny 1957). Moreover, Pezo- variance with all other findings, including ours, pot-us and Melopsittacus lack the characteristic that we reanalysed Ovenden’s (1984) raw elec- platycercine nape spot. The great genetic distance trophoretic data with the PAUP procedure. between Neophema bourkii and N, chrysostoma, Rooted by the mid-point method, it produced and their disparate separation in all algorithmic 17 equally parsimonious trees (length 45.0, con- analyses (Figs. 1, 2, 3a), implies that they are sistency index 0.821) none of which associated generically distinct. This result corroborates the Platycercus with Glossopsitta. From them a morphological and behavioral conclusions of “strict” consensus tree was derived and is de- Immelmann (1966) and Homberger (1980). picted in Figure 4. In it, Platycercus is aligned Two other lineages linked here to the Austra- with other broad-tailed parrots (Barnardius, Pse- lian platycercines are not corroborated in con- photus, etc.), albeit as a distinct sister lineage, temporary morphological studies. They are New while Glossopsitta is grouped with the cockatoos. Guinean Psittacella and the Australian polyteli- Obviously the single UPGMA phenogram of tine parrots, Alisterus and Polytelis. Psittacella Ovenden et al. (1987) doesnot representthe phy- has only been reviewed recently by Smith (1975) logenetic position of Platycercus accurately. who no doubt included it among psittaculine par- Among the remaining Psittacidae, other stud- rots because of its stout, short-tailed form and ies have singled out the same Australian platy- lack of a platycercine wing stripe and nape spot. cercine cluster as here, but with differing com- Nevertheless, Psittacella has the pale grey platy- position and internal relationships (cf. Table 1). cercinebill and the barred , colored rump There is consensus only over the inclusion of and blue cheeks (P. picta) found in some broad- Platycercus, Barnardius, Purpureicephalus, Nor- tailed parrots; its undertail coverts are also red, thiella and Psephotus. Moreover, our protein data contrasting with the belly as in Platycercus, Pur- corroborate the inclusion of the Swift Parrot (La- pureicephalus, Northiella and Psephotus. thamus) in this assemblage.The morphological The polytelitine parrots have been grouped similarities between Lathamus and the lorikeets, consistently with palaeotropic psittacine or psit- including nectar feeding, are evidently conver- taculine parrots by all current morphological re- gent (Smith 1975, Homberger 1980). views except those of Thompson (1899) and Several other genera conventionally included Verheyen (1956) which linked them, as here, to among the platycercines are placed ambiguously the platycercines. Smith (1975) and Homberger or excluded by our study. They are the Red- (1980) grouped the polytelitines with Geoffrovus, fronted Parakeet (Cyanoramphus), Ground Par- Eclectus, Psittacula and Agapornis in the tribe BIOCHEMICAL SYSTEMATICS OF PARROTS 315

Psittaculini. Our data corroborate a close rela- radiations. The first, which may date from the tionship between Geofroyus and Eclectus but onset of Tertiary aridity in mid Miocene times suggest,however, that the Psittaculini are poly- (Kemp 198 l), appears to have been continent- phyletic. This is reflected in chromosomal data wide and involved the divergence of the ground (Christidis et al., in press)which reveal that Agu- parrots (Pezoporus),Budgerigar (Melopsittacus), pornis, Psittacula and Alisterushave significantly Bourke’s Parrot (Neophema = Neopsephotus different karyotypes. The morphological and be- bourkii), typical Neophema, the core platycercine havioral characterssupporting the monophyly of group and the polytelitine parrots from one an- the Psittaculini need to be examined carefully to other. Outlying groups, such as ancestral stocks determine their nature, whether convergent, an- of Prosopeia, Cyanoramphus and cestral or derived. Certainly the phylogenetic sig- perhaps spread to Pacific islands over the same nificance of the carotenoid bill, which apparently period. The second radiation is centered in the unites the Psittaculini (Smith 1975) has to be eucalypt-dominated “Bassian” biota of southern re-assessedas it is also present in most lorikeets. Australia where the elements of the core platy- Biogeographicalandphylogenetic correlations. cercine group- Platycercus,Barnardius, Purpu- It is likely that the primary lineages of Australo- reicephalus,Northiella, Lathamus and Psepho- Papuan parrots and cockatoosarose on the Aus- tus-then split from one another. Members of tralian continental plate or were co-inherited from some of these genera have since adaptively col- (Cracraft 1973, Smith 1975, Rich onized the arid “Eyrean” and north Australian 1975) but without necessarily any “immigra- “Torresian” biotas (Barnardius, Northiella, Pse- tion.” Cockatoos, lorikeets and broad-tailed photus); but none has returned to the “Tumbu- platycercine parrots are the dominant groupsand nan” rainforests of the east coast. all three are centered today in the Australo-Pap- The genetic distances among the other pre- uan region. The cockatoos are probably one of sumed Australo-Papuan psittaculines- Eclec- the oldest lineages of the order. The genera of tus-Geofioyus, Cyclopsitta(fig-parrots) and Mi- lorikeets, in contrast, are little differentiated and cropsitta (pygmy-parrots)-and from their probably radiated recently. The three in Austra- supposedAfro-Asian relatives suggestthat they lia- , and Glossopsit- too many have arisen independently from as yet ta - have identical allelic constitutions in pro- unknown ancestral stocks in the Australian re- teins scored(Table 2). They could be considered gion. Their relationships are among the largest derivatives of a single lineage which entered Aus- gaps still to be resolved in the phylogeny of the tralia only recently from New Guinea. This is Psittaciformes. unlikely to have happened before Miocene-Oli- gocene times, 20-30 million years BP, because ACKNOWLEDGMENTS significant New Guinean land forms cut off from the Australian continental plate by the Aure For permissionto collectmaterial for this study,we are deeply indebtedto the Papua New Guinea De- trough had not developed before then (Doutch partmentof Environmentand Conservationand to the 1972, Dow 1977, Pieters 1982). AustralianState National Parksand Wildlife Services Concerning Psittacella, the occurrence of this in Victoria,, New SouthWales, South Aus- evident platycercine with ancestral morpholog- tralia, Tasmaniaand WesternAustralia, as well as the ConservationCommission of the NorthernTerritory. ical traits (type A-l carotid formula, barred Individual help wasgiven by K. Kisokau,L. Seri, G. plumage and no wing bar or nape spot) in the Kula, G. Maynes, P. Christian,A. Christidis,and I. montane rainforests of New Guinea is signifi- Schoddein PapuaNew Guinea, and by J. Boume,C. cant. These forestsand their vicariant biotas along Co&n, F. Crome,R. H. Green, M. King, L. Moore, the east coast of Australia hold a “Tumbunan” G. O’Neill, L. Pedler,J. C. Wombey, H. B. Gill, B. Gillies, J. Pedler,and P. Wilkinson in Australia.We avifauna that appears to comprise the surviving also acknowledgethe following for providing tissue elements of the Australian Tertiary rainforest and material:P. R. Baverstock(South Australian Museum) representatives of ancestral stocks from which for Melopsittacus undulatus. Micropsittapusio and Psi- many arid-adapted taxa have been derived tacella picta, A. Isles for Psittacula himalayani and P. derbiana and N. Hermes for Cyanoramphw novae- (Schodde and Calaby 1972). zelandiae. The figureswere preparedby F. Knight, CSI- The center of platycercine diversity is in Aus- RO Division of Wildlife and Ecology, Canberra, and tralia where protein data indicate two successive the manuscript was typed by T. Galariniotis and B. 316 L. CHRISTIDIS ET AL,.

Gillies, Museum of Victoria, Melbourne. Electropho- IMMELMANN,K. 1966. Die AustralischePlattschweif- resis was carried out in the Department of Population sittiche, 2nd ed. A Ziemsen, Wittemberg. Biology, Australian National University. The work was KEMP,E. M. 1981. Tertiary palaeogeographyand the funded by an ANU-CSIRO collaborativeresearch grant. evolution of Australian climate, p. 33-49. In A. Keast led.], Ecologicalbiogeography of Australia. Junk, The-Hague.- LI TERATURE CITED MAINARDI, D. 1962. Studio immunogenetico sulla posizione tassonomica di Melopsittacusundula- ADAMS, M., P. R. BAVERSTOCK,D. A. SAUNDERS, R. tus. Riv. Ital. Omitol. Ser. 2, 32:136-140. SCHODDE, AND G. T. SMITH. 1984. Biochemical NEI. M. 1978. Estimation of average heterozvnosity systematics of the Australian cockatoos (Psitta- ‘and genetic distance from a small number of in- ciformes: Cacatuinae).Aust. J. Zool. 32:363-377. dividuals. Genetics 89:583-590. AWSE, J. C., AND C. F. A&ADRO. 1982. A compar- OVENDEN, J. R. 1984. Investigationsinto the molec- ative summary of genetic distancesin the verte- ular evolution and phylogeny of Australian rosel- brates. Patterns and correlations. Evol. Biol. 15: las and related parrots. Ph.D. diss., Univ. New 151-185. South Wales. BEEHLER,B., ANDB. W. FINCH. 1985. Species-check- OVENDEN,J. R., MACK~NLAY,A. G., AND CROWER,R. list of the of New Guinea. AustralasianOr- H. 1987. Systematicsand mitochondrial genome nithol. Monogr. evolution of Australian rosellas (Aves: Platycer- BOET~ICHER,H. VON, 1964. Papageien. Wittenberg, cidae). Mol. Biol. Evol. 4:526-543. A. Ziemsen. PETERS, J. L. 1937. Checklist of birds of the world. BRERETON,J. LE GAY. 1963. Evolution within the Vol. 3. Harvard Univ. Press, Cambridge, MA. Psittaciformes, p. 499-517. In C.G. Sibley [ed.], Pnrrsas, P. E. 1982. Geology of New Guinea, p. 15- Proceedings 13th International Ornithological 38. In J.L. Gressit [ed.], Biogeographyand ecology Congress,American Ornithologist Union. Vol. 1. of New Guinea. Vol 1. Junk, The Hague. Ithaca, New York. RICH, P. V. 1975. Antarctic dispersal routes, wan- CHRISTIDIS,L., AND R. SCHODDE.In press. Relation- dering continents and the origin of Australasia’s ships of Australo-Papuan songbirds (Aves: Pas- non-passeriform avifauna. Mem. Nat. Mus. Vie. seriformes)-protein evidence. Ibis. 36:63-126. CHRISTIDIS,L., D. D. SHAW,AND R. SCHODDE.In press. ROOERS,J. S. 1972. Measures of genetic similarity Chromosomal evolution in parrots, lorikeets and and genetic distance. Studies in genetics7. Univ. cockatoos(Aves: Psittaciformes). Hereditas. Texas Publ. 7212:145-153. - CONDON,H. T. 1975. Checklist of the birds of Aus- ROHLF,F. J. 1982. Consensusindices for comparing tralia, Part 1. Non-passerines.Royal Australasian classifications.Math. Biosci. 59: 131-144. OrnithologistsUnion, Melbourne. SCHMUTZ,S. M., ANDS. E. PRUS. 1987. A cytogenetic CRAC~AFT,J. 1973. Continental drift, paleoclima- study of four speciesof cockatoos and Amazon tology,and the evolutionand biogeographyof birds. parrots. Genetica 74:69-7 1. J. Zool. 169:455-545. SCHODDE,R., AND J. H. CALABY. 1972. The bioge- DOUXH, H. F. 1972. The paleogeographyof north- ography of the Australo- Papuan and mam- em Australia and New Guinea and its relevance mal faunas in relation to Torres Strait, p. 257- to the Torres Strait area, p. l-10. In D. Walker 300. In D. Walker [ed.], Bridge and barrier: the [ed.], Bridge and barrier: the natural and cultural natural and cultural history of Torres Strait. Aus- history of Torres Strait. Anstralian National Univ. tralian National Univ. Press, Canberra. Press, Canberra. SHAW,C. R., AND R. PRASAD. 1970. Starch gel elec- Dow, D. B. 1977. A geological synthesis of Papua trophoresis-a compilation of recipes. Biochem. New Guinea. Australian Bur. Miner. Resour.Geol. Genet. 4:297-320. Geophys. Publ., Canberra. SIBLEY,C. G. 1960. The electrophoretic patterns of FARRIS,J. S. 1972. Estimatingphylogenetic trees from avian em-white nroteins as taxonomic characters. distance matrices. Am. Nat. 106645-668. Ibis lOz215-284. GLENNY,F. H. 1957. A revised classificationof the SIBLEY,C. G., AND J. E. AHLQ~I~T. 1972. A com- Psittaciformesbased on the carotid artery arrange- parative study of the egg-white proteins of non- ment patterns. Ann. Zool. (Agra) 2:47-56. passerinebirds. Bull. Peabody Museum Nat. Hist. GYSELS,H. 1964. A biochemical evidence for the 39:1-276. heterogeneityof the family Psittacidae. Sot. Roy. SMITH,G. A. 1975. Svstematicsof parrots. Ibis 117: Zool. Anvers. Bull. 33:2921. 18-68. HARRIS.H.. ANDD. A. HOPKINSON.1976. Handbook SNEATH.P.H.A.. ANDR. R. SOKAL. 1973. Numerical of enzymeelectrophoresis in human genetics.North . W.H. Freeman, San Francisco. Holland Publ., Amsterdam. SWOFFORD,D. L. 1981. On the utility of the distance HOLYOAK,D. T. 1973. Comments on taxonomy and Wagner procedure, p. 25-43. In V.A. Funk and relationshipsin the parrot subfamiliesNestorinae, D.R. Brooks feds.], Advances in cladistics. Pro- Loriinae and Platycercinae.Emu 73: 157-176. ceedingsof the F&t Meeting of the Willi Hennig HOMBERGER,D. G. 1980. Funktionell-Morpholo- Societv. New York Botanical Garden. Bronx. gischeUntersuchungen zur Radiation der Emahr- SWOFFORD:D. L. 1985. Phylogenetic analysis using ungs- und Trinkmethoden der Papageien(Psitta- parsimony (PAUP). Illinois Natural History Sur- ci). Bonn. Zool. Monogr. 13:1-192. vey, Champaign. BIOCHEMICAL SYSTEMATICS OF PARROTS 317

SWOFFORD,D. L., AND R. K. SEWER. 1981. A Chromosome studiesof 8 speciesof parrots of the computer program for the analysisof allelic vari- families Cacatuidae and Psittacidae (Aves: Psit- ation in genetics.J. Hered. 72:281-283. taciformes). Genetica 65: 109-l 17. THOMPSON, DA. W. 1899. On characteristicpoints VERHEYEN,R. 1956. Analyse du potentiel morpho- in cranial osteologyof the parrots. Proc. Zool. Sot. logique et projet dun nouvelle classificationdes London 1899:946. Psittaciformes. Bull. Inst. R. Sci. Nat. Belg. 32: l- VAN I)~NGEN, M.W.M., ANDL.E.M. DE BOER. 1984. 54.