sign in sign up
<<
Home , Margarornis, Ovenbird (family), Pearled treerunner, Premnoplex, Ruddy treerunner, Tyranni

The Condor945 ’60-166 0 The CooperOmithological Society 1992

THE PHYLOGENETIC RELATIONSHIPS OF THE ASSEMBLAGE (FURNARIIDAE) ’

DAVID Wiiss RUDGE AND ROBERT J. RAIKOW Department of Biological Sciences,University of Pittsburgh,Pittsburgh, PA 15260

Abstract. We performeda phylogeneticanalysis of the hindlimb musculaturein the Passerinegenera Margarornis, Premnornis,, and Roraimia (collectivelyknown as the Margarornis assemblageor “treerunners”)in orderto determinetheir affinitieswith one anotherand with the Dendrocolaptinae,a related,monophyletic group with similar scansorialhabits. We concludethat thetreerunners are not partof thedendrocolaptine , for they lack the synapomorphiesof that group.The treerunnersthemselves are shownto be monophyleticon the basisof five derivedcharacter states. Lochmias nematura, despite its outwardsimilarity to Margarornis squamiger,is excludedfrom thetreerunner assemblage by the absenceof the derivedcharacter states diagnosing that clade. Key words: Margaromisassemblage; treerunners; phylogeny; myologv; ; Lochmiasnematura.

INTRODUCTION (Ames 197 1; Feduccia 1973; Raikow, unpubl. REVIEW OF THE PROBLEM data) and on DNA-DNA hybridization (Sibley and Ahlquist 1985, 1990). They have been clas- The treerunners are a group of small, neotropical, sified by various workers either as separatefam- scansorial of the family Fumariidae, sub- ilies (Furnariidae and Dendrocolaptidae; e.g., family Fumariinae (Sibley and Monroe 1990). Vaurie 1980) or as subfamilies (Fumariinae and They include the Fulvous-dotted Treerunner Dendrocolaptinae) within the family Fumariidae (Margarornis stellatus), the (e.g., Sibley and Ahlquist 1990). We consider the (Margarornis squamiger), the two groups as subfamilies, following Sibley and (Margarornis rubiginosus), the Rusty-winged Monroe (1990). Barbtail (Premnornis guttuligera), the Spotted The present paper is part of an ongoing study Barbtail (Premnoplex brunnescens), and the Ro- of the relationships of the and wood- raimian Barbtail (Roraimia adusta). Margaror- creepers.In a previous paper (Rudge and Raikow nis bellulus and Premnoplex tatei were consid- 1992) we describedthe hindlimb myology of rep- ered by Vaurie (1980) as conspecific with resentatives of four treerunner genera in to Margarornis squamiger and Premnoplex brun- cataloguemuscle variations and to assessthe ex- nescens, respectively, but Sibley and Monroe tent to which they are correlated with the scan- (1990) list them as . In their scansorial sorial habit. Recent work on the appendicular habits, including some use of the tail as a prop, myology (Raikow, unpubl. data) has shown that the treerunners resemble the woodcreepers or the woodcreepers constitute a monophyletic Dendrocolaptinae, a consideration that prompt- ed Sclater (1890) to include them in his Den- group. drocolaptinae. In contrast, Feduccia (1973), fa- SPECIFIC AIMS voring family status for the ovenbirds and for We will addressthe following four questions. (1) the woodcreepers,grouped the treerunners with Are the treerunnerspart of the Dendrocolaptinae? the synallaxine fimariids, which he considered Because the Dendrocolaptinae is monophyletic the most distantly related to the Dendrocolap- (Raikow, unpubl. data), we must show that the tidae. treerunners either do or do not exhibit the syn- The ovenbirds and woodcreepers are widely apomorphies (shared derived character states) regardedas together constituting a monophyletic that diagnose the group. If the treerunners are group. This view is based both on morphology members of this clade, possessionof these char- acter states would corroborate this. (2) Are the treerunnersa monophyleticgroup?If the treerun- ’ ’ Received 10 February 1992. Accepted21 April ners constitutea clade, they should exhibit shared 1992. derived characterslacking in other fumariids. (3)

[7601 TREERUNNER RELATIONSHIPS 761

Is Lochmias nematura a member of the treerun- (Raikow, unpubl. data). None of the treerunners ner assemblage?Vatie (1980:238) proposedthis were found to possessthis character state. connection on the basis of plumage similarities (2) Origin of Mm. JIexor digitorum longus and with Margarornis squamiger. If Lochmias is a flexor hallucis longus. In the Dendrocolaptinae, member of the treerunner assemblage,it should M. flexor digitorum longus arises by the usual exhibit the synapomorphies of the group. (4) tibia1 and fibular heads, but also has a massive What are the relationshipsof the treerunnersto femoral head that occupiesa position in the crus one another?As with question (1) above, the an- usually held by the medial head of M. flexor swer will depend upon whether the treerunners hallucis longus (Fig. lA), which is correspond- are themselves monophyletic, plus the presence ingly reduced. This condition is derived for the of derived characterstates within the assemblage woodcreepers (Raikow, unpubl. data). In tree- that diagnose subordinate . runners, however, M. flexor digitorum longus lacks a femoral head (Fig. lB), and the medial MATERIALS AND METHODS head of M. flexor hallucis longus is not reduced. Twenty-five specimens of ten species of birds This is the primitive passerinearrangement, and preservedin ethanol were dissected.Sample sizes fails to support the inclusion of the treerunners and catalog numbers are given below in paren- in the dendrocolaptine clade. theses, with abbreviations for museums from (3) Tendon ossification.The Dendrocolaptinae which specimens were borrowed as follows: have evolved ossification of most of the crural AMNH, American Museum of Natural History; tendons, in contrast to the usual con- CM, Carnegie Museum of Natural History; FM, dition in which most of these tendons are not Field Museum of Natural History; LSU, Loui- ossified. The treerunners lack ossified tendons siana State University Museum of Natural Sci- except for the tarsal segment of the tendon of M. ence:Margarornis rubiginosus(2: LSU/JPO 2289, flexor digitorum longus, which alone is some- LSU 64786); Margarornis squamiger (5: LSU times found in other .The absence of 60811, LSU 75008, LSU 79559, LSU 79560, this complex derived state argues against a LSU 33 1197); Premnoplex brunnescens(7: FM grouping of the treerunners with the Dendro- 321530, FM 321531, LSU 70631, LSU 75010, colaptinae. LSU 85962, LSU 89447, LSU 95543); Prem- (4) Insertion of M. pubo-ischio-femoralispars noris guttuligeria (5: LSU 83865, LSU 89444, caudalis. In the Dendrocolaptinae, M. pubo-is- LSU 89445, LSU 107624, LSU 107625); Ro- chio-femoralis pars caudalis inserts onto the fe- raimia ad&a (1: AMNH 7685); Lochmias ne- mur alone, a derived state, and not additionally matura (1: LSU 79570); Dendrocolaptescerthia upon M. gastrocnemius pars intermedia as in (1: CM 1380); Deconychura longicauda (1: LSU other passerine birds (Raikow, unpubl. data). 1076 17); anabatina (1: CM 38 12); The treerunners exhibit the typical passerine Deconychurastictolaema (1: LSU 114403). Dis- condition (Rudge and Raikow 1992, Figs. 7, 8) section methods were described previously and do not exhibit the dendrocolaptine state for (Rudge and Raikow 1992). this character. Abbreviations for Figures l-5 are as follows: (5) Insertion of M. flexor cruris medialis. Rai- FCRLA, M. flexor cruris lateralis pars accessoria; kow (unpubl. manuscript) pointed out that the FCRLP, M. flexor cruris lateralis pars pelvica; Dendrocolaptinae exhibit changesin the size and FIB, head of the fibula; FL, M. fibularis longus; insertion of the M. flexor cruris medialis tendon FPD4, M. flexor perforatus digiti IV, FPPD2, M. relative to the insertion of Tendon M. None of flexor perforans et perforatus digiti II; FPPD3, the treerunners exhibited such changes (Rudge M. flexor perforans et perforatus digiti III; G, M. and Raikow 1992, Fig. 8). gastrocnemius; IF, M. iliofibularis. (6) Curvature of the claws. Raikow (unpubl. manuscript) reported that the claws of digits II- RESULTS AND DISCUSSION IV in woodcreepershave acquired a greater arc of curvature than those of ovenbirds. In the tree- RELATIONSHIP TO THE DENDROCOLAPTINAE runners, however, the arcs of the claws remain (I) Digit length. In dendrocolaptines digit IV is within the range of nonscansorial ovenbirds, and elongated nearly to the length of digit III, a de- lack the derived specialization of the woodcreep- rived state in contrast to most passerine birds ers. 762 DAVID W. RUDGE and ROBERT J. RAIKOW

Medial head

dial head

3 mm FIGURE 1. Comparison of the origin of M. flexor digitorum longus between: (A) Dendrocolaptescerthia CM 1380 (Dendrocolaptinae); and (B) Premnoplex brunnescensLSU 70631 (Furnariinae), illustrating the absence of a femoral head in the latter. Abbreviations for the figuresare given under Materials and Methods.

(7) Additional characters. Other charactershave support inclusion of the treerunners in the den- been used to separate woodcreepersfrom other drocolaptine clade. passerines.“ Horns” on the (Ames 1971) arguefor monophyly. In ovenbirds, “horns” were found only in (Ames 197 1: 154) how- MONOPHYLY OF THE TREERUNNERS ever Ames (197 1) apparently did not examine In order to demonstrate the monophyly of the the syringesof the treerunners. Dendrocolaptines treerunner assemblage, it is necessary to show also characteristically possessa stiffened tail, but that its members share synapomorphies lacking Vaurie (1980:230) indicated that the tail of tree- in other forms. We will consider whether any of runners is not notably stiffened. the muscle variations found in our dissectionsis In summary, none of the synapomorphic hind- such a synapomorphy, using the method of out- limb character states that delimit the Dendro- group comparison (Maddison et al. 1984, Rai- colaptinae was found in any of the treerunner kow 1987:7-S). This method requires that the specimens, nor did the traditional characters speciesin question belong to some more inclu- TREERUNNER RELATIONSHIPS 763 sive clade (Raikow 1982:433). Raikow (unpubl. pattern of its occurrence suggeststhat it is due manuscript) has established that the - to individual variation. Therefore, we conclude assemblageis probably monophy- that it has probably been independently derived letic. A rigorous application of outgroup com- in the treerunner assemblage and is a synapo- parison would demand a knowledgeof the precise morphy of the group. sister-group relationships of the treerunner as- (3) Reduction of M. flexor perforatus digiti IV. semblage. This is not available at the present In suboscines, M. flexor perforatus digiti IV is time, but Raikow (unpubl. manuscript) dissected composed of two separate bellies (Pars proxi- a representative sample of the Furnariinae, pro- malis and Pars distalis), each of which gives rise viding a basis for comparison. We used succes- to a tendon that joins with the other to form a sive sister groups of the ovenbird-woodcreeper common tendon of insertion. In treerunners, Pars assemblageas outgroups (more distantly related proximalis is unusually small, and its insertion Furnarii, including Thamnophilus doliatus, For- is reduced to a slender, hair-thin tendon (Fig. micivora mental& Cercomacra tyrannina, and 3A). This condition is unknown in other mem- Acropternisorthonyx; New World ; and bers of the fumariine-dendrocolaptine assem- Old World suboscines). We also made note of blage (Fig. 3B) (Raikow, unpubl. data). The ab- the condition of the character in question in the sence of this character state in other suboscines Dendrocolaptinae. These groupingsrepresent the (Raikow, unpubl. manuscript; McKitrick 1985: best current estimate of passerine phylogeny 296; Raikow 1987:22-23) suggeststhat it is a (Raikow 1987, Bledsoeand Raikow 1990, Sibley derived condition in the Fumariidae, and there- and Ahlquist 1990). fore synapomorphic for the treerunner assem- (I) Origin of M. flexor cruris medialis. In most blage. passerines,the caudal extent of the origin of M. (4) M. flexor hallucis longus. A swelling oc- flexor cruris medialis arises from the edge of the curred in the insertion tendon of M. flexor hal- dorsal rim of the ischiopubic fenestra;but in tree- lucis longus (Rudge and Raikow 1992, Fig. 9) in runners the origin extends from beyond the rim, all specimens. This condition is lacking in other forming a characteristic‘ lip’ (Rudge and Raikow members of the fumariine-dendrocolaptine as- 1992, Fig. 5). Despite some variation in size, this semblage (Raikow, unpubl. manuscript). It has occurred in all specimens. Such an origin is un- not been reported in other suboscines(McKitrick known in other members of the fumariine-den- 1985:304, Raikow 1987:23-25), suggestingthat drocolaptine group (Raikow, unpubl. data). Its it is a derived condition in the Fumariidae and apparent absence in the other Fumarii men- synapomorphic for the treerunner assemblage. tioned above (Raikow, unpubl. manuscript), the (5) M. lumbricalis. This is a small muscle that New World Tyranni (McKitrick 1985:289; Rai- extends from the trifurcation of the M. flexor kow, unpubl. observ.), and the Old World subos- digitorum longus tendon to the joint pulleys of tines (Raikow 1987:14) suggeststhat this is a the forward toes. In treerunners, one or two ad- derived condition in the Fumariidae and, thus, ditional bellies were found in all specimens synapomorphic for the treerunner assemblage. (Rudge and Raikow 1992, Fig. 9). This is un- (2) Origin of M. flexor perforans et perforatus known in other members of the fumariine-den- digiti II. In treerunners the origin of M. flexor drocolaptine assemblageor other Fumarii (Rai- perforans et perforatus digiti II extends from the kow, unpubl. manuscript), or the Old World laterodistal surfaceof the femur (as in other pas- suboscines(Raikow 1987:27). M. lumbricalis is serines)caudolaterally along the lateral collateral absent among the New World Tyranni (Mc- ligament, which joins the femur to the head of Kitrick 1985:306). Although the presenceoftwo the fibula (Rudge and Raikow 1992, Fig. 8). The bellies has been reported in some nonpasserine absenceof such an extended origin among other birds (i.e., Gallicolumba luzonica and Opistho- members of the fumariine-dendrocolaptine as- comushoazin: George and Berger 1966:464), the semblage (Fig. 2) and other suboscines(Raikow, absence in passerinesof an association with the unpubl. manuscript; McKitrick 1985:294; Rai- flexor hallucis longus tendon (found in those kow 1987:21), demonstratesthat this is a derived nonpasserineshaving a secondbelly) suggeststhat state among fumariines. It has also been found this condition has been independently derived unilaterally or bilaterally in a few dendrocolap- and, as such,is a synapomorphy of the treerunner tines (Raikow, unpubl. manuscript), where the assemblage. 764 DAVID W. RUDGE and ROBERT J. RAIKOW

FPPD2

F6PD3 FL‘ 3mm

FIGURE 2. Origin of M. flexor perforanset perforatusdigiti II (FPPD2) in the dendrocolaptineDeconychura stictoluemaLSU 114403, showing the absenceof an origin from the head of the fibula (FIB).

Thus, our dissectionssuggest five synapomor- RELATIONSHIP OF LOCHMZAS TO THE phic charactersthat appear to demarcate a tree- TREERUNNER ASSEMBLAGE runner clade from other members of the oven- Lochmias nematura has been suspectedof affin- -woodcreeper assemblage. ities with Margarornis squamiger on the basis of

Pars proximalis

\ Pars distalis

FIGURE 3. Comparison of M. flexor perforatus digiti IV in (A) Margurornis squamigerLSU 75008; and (B) Dendrocoluptescerthiu CM 1380. Note the reduction of Pars proximalis and its fine, hairlike tendon in the treerunners compared to its condition in the Dendrocolaptinae. TREERUNNER RELATIONSHIPS 165 a similar pearly plumage pattern and similarities Roraimia adusta in the tail (Vaurie 1980:238). If Lochmias is a member of the treerunner assemblage,it should Margarornis squamiger exhibit the synapomorphies of that group. In Lochmias M. flexor cruris medialis doesnot have Margarornis rubiginosus a caudally extended origin as in the treerunner assemblage;its origin extends only to the rim of Premnornis guttuligera the ischiopubic fenestra. M. lumbricalis consists of only one belly in Lochmias. M. flexor perfor- Premnoplex brunnescens ans et perforatus digiti II in Lochmias lacks an FIGURE 4. illustrating relation- extended origin onto the head of the fibula, but shipsof the Margarornisassemblage based upon hind- retains the typical origin found in other subos- limb muscle synapomorphies:(1) M. flexor cruris me- dialis origin extended;M. flexor perforanset perforatus tines. M. flexor perforatus digiti IV in Lochmias digiti II origin extended; M. flexor perforatus digiti IV has neither the reducedbelly nor the fine, hairlike pa& proximalis belly and tendon reduced; swelling in tendon characteristic of the treerunner assem- tendon of M. flexor hallucis longus: M. lumbricalis blage. Lochmias does have a slight stiffening in possessionof two bellies; (2) M.-ilibtibialis lateralis parspostacetabularis belly reduced;(3) Tendon M orig- the region of the swelling of M. flexor hallucis ination from the tibia1lobe of M. flexor cruris lateralis longus noted in the description of this muscle pars pelvica; M. lumbricalis possessionof third inter- above, but no swelling is present. We interpret mediate belly. this as not being sufficiently similar to constitute a character state shared with the treerunner as- semblage. In conclusion, Lochmias has none of the synapomorphies of the treerunners, and we rived for the clade consisting of Premnoplex therefore conclude that it is not a member of that brunnescensand Premnornis guttuligera. group. One of the more striking differencesamong the speciesof the present study is the presence of a INTERRELATIONSHIPS OF THE third belly of M. lumbricalis (Rudge and Raikow TREERUNNERS 1992, Fig. 9) in Premnornis guttuligera and The caudal extent of the postacetabular belly of Premnoplex brunnescens.Some individual ex- M. iliotibialis lateralis was reduced in all Mar- ceptions were found in both these species, but garornis specimens,in contrast to the other spe- this may be due to dissection artifact, given the cies of this study (Rudge and Raikow 1992, Figs. tiny size of the structure. No other birds in this 1, 2). Such reductional trends have been ob- study had a third belly, which is also unknown served in a number of passerine groups (e.g., in other birds (Georgeand Berger 1966:483-484). M&trick 1985:308,Raikow 1987:28)including We conclude that this condition representsa syn- a few furnariines and dendrocolaptines (Raikow, apomorphy for these two species. unpubl. manuscript). This outgroup comparison Our conclusions are presented in the form of is therefore somewhat equivocal. However, as a tree (Fig. 4). Information from the hindlimb most outgroupseither lack this character,or have musculature is insufficient to completely resolve it in some but not all species, it probably rep- the phylogeny of the treerunners, but it does pro- resentsa derived state in the treenmners as well, vide a documented hypothesis for future com- and constitutes a synapomorphy for the two parison with other studies. Margarornis species. ACKNOWLEDGMENTS In Premnoplex brunnescensand Premnornis We thank William P. Coffman, Harry 0. Corwin, and g-uttuligera,Tendon M originates entirely from Frederick J. Gottlieb of the Department of Biological the tibia1 lobe of M. flexor cruris lateralis pars Sciences,University of Pittsburgh,for assistancein the pelvica (Rudge and Raikow 1992, Figs. 7, 8). A planning and execution of this study. We are particu- similar modification has also been found among larly grateful to Anthony H. Bledsoe and Mary C. dendrocolaptines (Raikow, unpubl. manuscript). M&trick for many helpful contributions. We are also grateful to the following members of the This condition is lacking in other Fumarii and Carnegie Museum of Natural History Section of Birds in the more distant New World Tyranni, and, for their help and advice during the project: Kenneth given the monophyly of the treerunners, is de- C. Parkes, D. Scott Wood, and James Loughlin, who 766 DAVID W. RUDGE and ROBERT J. RAIKOW provided specimens,advice, and encouragement,and RAIKOW, R. J. 1982. Monophyly of the passeri- Robin Panza. who drew the fiaures. We thank Scott formes: test of a phylogenetichypothesis. Auk 99: Lanyon (Field Museum of Natural History), J. Van 431-445. Remsen (Louisiana State University Museum of Nat- RAIKOW, R. J. 1987. Hindlimb myology and evolu- ural Science), and George Barrowclough (American tion of the Old World suboscinepasserine birds Museum of Natural History) for loans of specimens, (Acanthisittidae, Pittidae, Philepittidae, Eury- and the Department of Biological Sciences,University laimidae). Omithol. Monogr. No. 41. American of Pittsburgh, for providing funding for the illustra- Ornithologists’ Union, Washington, DC. tions. RUDGE,D. W., AND R. J. RAIKOW. 1992. Structure, This material is based upon work supported by the function, and variation in the hindlimb muscles National Science Foundation under Grant No. BSR- of the Murgurornis assemblage (Aves: Passeri- 8617896 to R. J. Raikow. formes: Fumariidae). Ann. CarnegieMus. 61:207- 237. SCLATER,P. L. 1890. Catalogue of the birds in the LITERATURE CITED British Museum, Vol. 15. Brit. Mus. Nat. Hist., London. AMES,P. L. 1971. The morphology of the syrinx in SIBLEY,C. G., AND J. E. AHLQUIST. 1985. The phy- uasserinebirds. Bull. Peabodv Mus. Nat. Hist. 37: logeny and classificationof the New World subos- i-194. tine passerinebirds (Passeriformes:Oligomyodi: BLEDSOEA. H., AND R. J. RAIKOW. 1990. A quan- ), p. 396428. In P. A. Buckley, M. S. titative assessmentof congruencebetween molec- Foster, E. S. Morton, R. S. Ridgely, and F. G. ular and nonmolecular estimates of phylogeny. J. Buckley [eds.], Neotropical ornithology. Omithol. Mol. Evol. 30~247-259. Monogr. No. 36. American Omitholoaists’ Union. FEDUCCIA,A. 1973. Evolutionary trends in the neo- Washington, DC. tropical ovenbirds and woodhewers. Omithol. SIBLEY,C. G., ANDJ. E. AHLQUIST. 1990. Phylogeny Monogr. 13:1-69. and classificationof birds. Yale Univ. Press,New GEORGE,J. C., AND A. J. BERGER. 1966. Avian my- Haven, CT. ology. Academic Press, New York. SIBLEY,C. G., AND B. L. MONROE,JR. 1990. Distri- MADDISON,W. P., M. J. DONOGHUE,AND D. R. MAD- bution and of birds of the world. Yale DISON. 1984. Outgroup analysisand parsimony. Univ. Press, New Haven, CT. Syst. Z.ool. 33:83-103. VAURIE, C. 1980. Taxonomy and geographicdistri- MCKITRXZK,M. C. 1985. Pelvic myology of the king- bution of the Fumariidae (Aves. Passeriformes). birds and their allies (Aves: Tyrannidae). Ann. Bull. Am. Mus. Nat. Hist. 166:ll375. ’ Carnegie Mus. 54:275-317.