SUMMARY The degree of toe fusion, here summarized for the oscine families, probably has only limited taxonomic usefulness, but may serve along with other characters to detect genera possibly needing further systematic study, e.g., Chlamydochaera, Ifrita, Newtonia. Terrestrial species often have lower fusion than do arboreal or climbing species, but there are important exceptions. Low fusion apparently aids balance on flat substrates, whereas high fusion facilitates perching or climbing. The large number of arboreal species with low fusion has not yet been satisfactorily explained.

Although striking differences in extent of webbing between the toes of nonpasserines are well-known and often used to illustrate adaptation in birds, the degree of connection between the toes in oscines has received much less attention. Ridgway (1901-07) in scattered keys and descriptions commented on integumental fusion of the toes of numerous New World oscine taxa, but provided little interpretation on possible significance of the taxonomic variations, apart from their use in distinguishing taxa. Al- though others (e.g., Rand and Traylor 1953) have occasionally commented on fusion in oscines, a comprehensive survey of the families is lacking. ! attempt here to interpret major taxonomic differences in fusion in relation to systematics and behavioral differences and to indicate problems for future study. MATERIALS AND METHODS I examined study skins of 1941 species of oscines in the collections of the National Museum of Natural History (Washington, D.C.), American Museum of Natural History (New York), British Museum (Nat. Hist., Tring) and the University of Connecticut. In addition, I made more than 175 observations of toe positions in perching or standing for wild or captive birds representing 30 species. The taxonomic sequence follows Morony, Bock and Farrand (1975). To examine fusion among the 3 forward toes I used a hand lens or binocular dissecting microscope, except for species of large size. Degree of fusion of the middle (III) and outer (IV) toes is primarily emphasized, but extent of fusion of the inner (II) and middle toes was noted for 349 species, as discussed below. I selected as a major landmark for comparison the articulation between the first and second phalanges of the middle toe, located by bends in that toe, by the plantar flexion creases and, in many cases, by an overlying scute termed the proximal cap (Clark 1977). Fusion of toes III and IV is rated low (L) if not reaching distally to the region of articulation between the first and second phalanges of III, moderate (M) if reaching that region, and high (H) if extending further distally (Fig. 1). These ratings are arbitrary divisions of a continuum, but the extremes of low vs high fusion are markedly different. This summary of fusion in 3 categories provides less detail than sometimes given by Ridgway (1901-07), but is advantageous in facilitating comparisons across a wide range of taxa. RESULTS AND INTERPRETATION In tabulating data on fusion of toes III and IV (Table 1), I emphasize particularly the distribution of extreme differences (low vs high) and the condition in many genera considered atypical in their assigned families. Such a summarizing list (Table 1) necessarily obscures many finer taxo- nomic differences. For example, nearly all genera of the Mimidae have 67 68 THE WILSON BULLETIN ß Vol. 93, No. 1, March 1981 FIG. 1. Examples of toe fusion, from left to right: low, House Sparrow (Passer domesti- cus); moderate, White-breasted Nuthatch (Sitta carolinensis); and high, Red-eyed Vireo (Vireo olivaceus). The arrows mark the approximate level of the joint between phalanges 1 and 2 of the middle toe. low fusion, but the inclusion of Donacobius extends the mimid range to the moderate level. Partial associations exist between degree of fusion and behavior. The terrestrial larks (Alaudidae) and pipits (Motacillidae) have low to moderate fusion, whereas many predominantly arboreal Old World families have moderate to high fusion, e.g., Campephagidae, Irenidae, Dicaeidae, Nec- tariniidae, Zosteropidae, Meliphagidae, Oriolidae, Dicruridae. Families that climb on tree trunks or other vertical surfaces often have moderate to high fusion, e.g., Sittidae, Certhiidae, Climacteridae. Still other Old World families or subfamilies range from low to high fusion, e.g., Laniidae, Timaliinae, Sylviinae, Malurinae. Finches, sparrows and buntings throughout the world (Emberizinae, Cardinalinae, Carduelinae, Estrildi- dae, Ploceidae) have low to moderate fusion, a level widespread in the New World 9-primaried assemblage, including Parulidae, Drepanididae and Icteridae. However, vireos (Vireonidae), including peppershrikes (Cy- clarhis) and shrike-vireos (Vireolanius), have greater fusion. Species within a genus are usually similar in the broad categories of toe fusion used here (Table 1), but a few Old World genera, the warblers Cettia and Bradypterus and the bush-shrikes Telophorus, exhibit excep- tional interspecific differences. Within Cettia, for example, the low fusion of C. squameiceps, C. major and C. brunnifrons contrasts with high fusion of C. fortipes; other species are intermediate. Among the 10 examined species of Bradypterus, only B. seebohmi has high fusion, the other 9 being low. Telophorus bocagei, T. sulfureopectus, T. olivaceus, T. nigri- frons and T. multicolor have high fusion in contrast to low to moderate in T. zeylonus, T. viridis, T. quadricolor and T. dohertyi. In Telophorus, separation of groups of species by degree of fusion matches taxonomic units recognized on other characters (Hall and Moreau 1970), but fusion differences within Cettia and Bradypterus do not parallel taxonomic Clark, Jr. ß OSCINE TOE FUSION 69 TABLE 1 FUSION OF TOES III AND IV IN OSCINES Taxa No. species examined Fusion Alaudidae 39 L-M Hirundinidae 43 L-M plus Atticora 2 H Neochelidon 1 H Motacillidae 47 L Campephagidae 53 M-H plus Chlamydochaera 1 L Pycnonotidae 12 M-H plus Spizixos 1 L H ypsipetes 1 L Irenidae 4 M-H Laniidae Prionopinae 7 M-H Malaconotinae 24 M-H plus Telophorus 9 L-H Laniinae 22 L-M Pityriasinae 1 H Vangidae 9 H plus Hypositta 1 H Bombycillidae 6 L plus Phainoptila 1 M Dulidae 1 M Cinclidae 5 L Troglodytidae 20 L-M Mimidae 12 L-M Prunellidae 8 L Muscicapidae Turdinae 261 L-M Orthonychinae Orthonyx 2 H Androphobus 1 H Psophodes 1 M Sphenostoma 1 M Cinclosoma 1 L Eupetes 1 M Melampitta 1 L lfiita 1 H 70 THE WILSON BULLETIN ß Vol. 93, No. 1, March 1981 TABLE 1 CONTINUED Taxa No. species examined Fusion a Timaliinae 38 L-M plus Garritornis 1 H Stachyris 10 M-H Rhopocichla i H Macronus 4 M-H Micromacronus I H Timalia I H Pteruthius 4 M-H Alcippe 7 M-H Yuhina 3 M-H Panurinae 7 L-M Picathartinae I L Polioptilinae Microbates 1 H Ramphocaenus I H Polioptila 2 M Sylviinae 128 L-M plus Psamathia I H Cettia 9 L-H Bradypterus 10 L-H A crocep halus 21 M-H Hippolais 3 M-H Bathmocercus I H Macrosphenus 1 H Malurinae Malurini 12 M-H plus Amytornis 2 L Stipiturus I L Acanthizini 26 L-M plus Gerygone 5 M-H Mohouini 3 M-H Epthianurini 4 L-M Genus incertae sedis Lamprolia 1 M Muscicapinae Bradornis 2 L-M Melaenornis 3 L-M Fraseria 1 M Rhinomyias 2 L-M Ficedula 15 L-M Niltara 12 L-M Clark, Jr. ß OSCINE TOE FUSION 71 TABLE 1 CONTINUED Taxa No. species examined Fusion a Muscicapa 14 L-M Myioparus 2 L-M Humblotia 1 M Newtonia 2 M-H Microeca 4 M-H Peltops 2 M-H Petroica 5 L-M Tregellasia 3 M Eopsaltria 2 M Philentoma 1 H Poecilodryas 4 M-H Peneothello 1 M Pach ycephalopsis 1 H Platysteirinae 16 M-H Monarchinae 54 M-H Rhipidurinae 22 M-H Pachycephalinae 34 M-H plus Hylocitrea 1 L Genus incertae sedis Turnagra 1 M Aegithalidae 7 M-H Remizidae 7 M-H Paridac 33 M Sittidae 22 M-H C erthiidae 6 M-H Rhabdornithidae 2 M Climacteridae 5 H Dicaeidae 48 M-H Nectariniidae 94 M-H Zosteropidae 47 M-H Meliphagidae 110 M-H Emberizidae Emberizinae 42 L-M Catamblyrhynchinae 1 M Cardinalinae 8 L Thraupinae 24 L-M Tersininae 1 L Parulidae 41 L plus Zeledonia 1 L Drepanididae 11 L-M Vireonidae 38 M-H Icteridae 19 L-M 72 THE WILSON BULLETIN ß Vol. 93, No. 1, March 1981 TABLE 1 CONTINUED Taxa No. species examined Fusion a Fringillidae Fringillinae 1 L Carduelinae 17 L-M Estrildidae 9 L-M Genus incertae sedis Pholidornis 1 M Ploceidae 15 L-M Sturnidae 14 L plus Buphagus 1 M Oriolidae 22 M-H Dicruridae 17 M-H Callaeidae 3 L Grallinidae 4 M Artamidae 9 M-H Cracticidae 7 M-H Ptilonorhynchidae 11 L-H Paradisaeidae 31 M-H Corvidae 93 L-M plus Platylophus 1 H Crypsirina 2 M-H Temnurus 1 H a Symbols: L = low, M = moderate, H: high. groups of species. Unfortunately, little has been reported about the use of the feet in Cettia, Bradypterus and Telophorus; species of the first 2 are widely noted as difficult to observe as they skulk in brush. Fusion of toes H and III.-Fusion of toes II and III extends less far distally than that between III and IV. The relatively few taxa with fusion of II and III reaching distally to the vicinity of the articulation of phalanges 1 and 2 of toe III also have high fusion of toes III and IV, e.g., Vangidae, Orthonyx, Microbates, Ramphocaenus, certain muscicapid flycatchers, Hypositta, Certhia, Climacteris, Vireonidae. Evolution and systematics.--The taxonomic distribution of different de- grees of fusion including variation within genera, subfamilies and families shows that evolutionary convergence has been frequent. Among birds as a whole, and among oscines, high fusion between toes is probably usually a derived, rather than primitive, condition. However, reduction of high Clark, Jr. ß OSCINE TOE FUSION 73 fusion remains a hypothetical possibility and might have occurred occa- sionally. Raikow (1978) suggested that in situations where evolutionarily primitive and derived conditions are indeterminable, systematists should use characters phenetically while recognizing that resulting hypotheses on relationships will be relatively weak. Any systematic suggestions for os- cines based heavily on similarity of toe fusion would be at best tentative, but where fusion agrees with other characters in differing markedly be- tween genera traditionally hypothesized to be closely related, reconsider- ation of affinities seems warranted, as in certain of the following examples. The monotypic Chlamydochaera from Borneo is the sole genus of the cuckoo-shrikes (Campephagidae) with low fusion. Ames (1975) found that Chlamydochaera was unique among examined campephagids in having a thrush-like syrinx and concluded that the genus belongs in the thrushes (Turdidae). The low fusion of the toes is also like that of thrushes. In addition, Ames found thrush-like syringes in the muscicapine genera Bra- dornis, Melaenornis, Rhinomyias, Ficedula, Niltara and Muscicapa, which, unlike most other Old World flycatchers, also have relatively low toe fusion like that of thrushes. Harrison (1967) suggested that the babbler Ifrita from New Guinea is closely related to the blue wren group of Clytomyias, Chenorhamphus, Todopsis and Malurus of Australia and New Guinea. I find that all these genera share a high fusion of toes III and IV. In addition, Ifrita has a ridged culmen like that of Clytomyias and shares the unusual feature of blue feathering on the head with Todopsis, Malurus and male Chenorham- phus. Thus, several characters link Ifrita with the malurids, indicating a possible relationship not reflected in traditional classification. Clytomyias, Chenorhamphus, Todopsis and Malurus have an unusual gap in the inter- scapular zone of the spinal feather tract, present also in the Australian grass-wrens (Amytornis) and emu-wrens (Stipiturus; Harrison 1969); the condition of Ifrita in this regard has not been reported. As the relatively terrestrial Amytornis and Stipiturus have low fusion, caution is necessary in using fusion as a taxonomic character in this group. Dorst (1960) proposed that the monotypic genera Tylas and Hypositta from Madagascar belong in the family Vangidae, which is endemic to that island. The high fusion in all these birds is compatible with his suggestion, although high fusion also occurs in other families with which Tylas has often been placed, including bulbuls (Pycnonotidae). In Hypositta, toe fusion is greater than in typical nuthatches (Ridgway 1904:439; this study), and Hypositta is thus more like the vangids in this respect. Also from Madagascar is Newtonia, the sole genus of the Muscicapinae outside the Australian region having moderate to high fusion. Although superficial appearances of study skins can be highly misleading concerning evolu- 74 THE WILSON BULLETIN ß Vol. 93, No. 1, March 1981 tionary affinities, Newtonia brunneicauda and females of the vangid Cal- icalicus madagascariensis have some resemblance. Further consideration of the affinities of Newtonia would be desirable. The high fusion of Cyclarhis, Vireolanius, Vireo and Hylophilus sup- ports the idea that these genera constitute a monophyletic group (Barlow and James 1975, Raikow 1978). Such high fusion is unusual among New World oscines, being known otherwise from Certhia and 2 genera of swal- lows, Microbates, Ramphocaenus. Vireos thus differ markedly from most New World 9-primaried oscines, including warblers, tanagers, blackbirds and buntings. Use of the feet.--As relationships between the degree of fusion and use of the feet are not well understood, ! have emphasized here the most conspicuous taxonomic differences, for associations between structure and behavior might be most prominent in such cases. My findings support Riiggeberg's (1960) conclusions, based on a much smaller sample of species, that high fusion often occurs in arboreal species and that low fusion is typical for terrestrial species. Bock and Miller (1959) indicated that the high fusion of syndactyly in nonpasserines was advantageous in arboreal perching because the parallel position of the toes applies all the force of flexion directly against a branch; separated toes would presumably be mechanically less efficient. In those climbing oscines with syndactyly, the forward toes are restrained in a roughly parallel orientation that pos- sibly helps to ensure a secure grasp on vertical surfaces. My observations of live oscines and of published photographs indicate that birds with low fusion vary the spread of the forward toes considerably according to the kind of perch. On the ground or other flat surfaces, these toes are widely separated, presumably providing a stable base for standing or moving. However, on horizontal perches of a small diameter relative to foot size the forward toes are held close together, a position equivalent to that of syndactyly, with presumably similar advantages. On sharply in- clined perches of small diameter, toe I! of the lower foot is often abducted from III and IV, which are held close together (Leisler 1972; see also Willis 1969, 1972); the application of forces in 2 directions against the perch by toe II, as opposed to III and IV, presumably helps to prevent the foot from slipping down the perch. Leisler (1972) has provided further details on the relationships between toe position and body orientation of small oscines perched on vertical stems. A lack of absolute associations between degree of syndactyly and use of the feet prevents the use of structure to predict habits of species not studied alive. For example, the climbing Black-and-white Warbler (Mni- otilta varia) has low fusion, like that of allied nonclimbing parulids (Parkes 1978; this study), in contrast to the moderate to high fusion of many trunk- Clark, Jr. ß OSCINE TOE FUSION 75 climbers in other families. The tree creepers (Certhia), which have highly fused forward toes and stiflened tail feathers, are climbers. Because Or- thonyx from Australia and New Guinea have the same structural charac- teristics they might be expected also to be climbers; however, Orthonyx actually forage terrestrially, propping themselves on the ground with 1 leg and stiffened tail while scratching in the litter with the other foot (Zusi 1978). As another example of absence of absolute associations between toe fusion and degree of arboreality, many arboreal species have only low to moderate fusion, e.g., kinglets (Regulus) and New World orioles (Ic- tertts). Possibly such arboreal birds with low fusion differ in perching hab- its from those with high fusion, but evidence is lacking. Although asynchronous terrestrial gaits (walking and running) are char- acteristic for terrestrial oscines, and synchronous (hopping) for arboreal species, there are many exceptions (Clark 1975), and degree of toe fusion is not absolutely associated with gait, except that walking is apparently exceptional in oscines with high fusion. Furthermore, no direct association exists between degree of fusion and the ability to hold food with the feet (Clark 1973). Relationships between fusion and uses of the feet may not be apparent in tnany cases without simultaneously considering many other aspects of structure and behavior. ACKNOWLEDGMENTS I thank G. E. Watson, J. Farrand, Jr. and I. C. J. Galbraith for providing access to specimens in their care. R. L. Zusi gave helpful advice and J. A. Slater made useful sug- gestions on an earlier version of the manuscript. Miss Mary Hubbard prepared the illustra- tion. LITERATURE CITED AMES, P. L. 1975. The application of syringeal morphology to the classification of the Old World insect eaters (Muscicapidae). Bonn. Zool. Beitr. 26:107-134. BARLOW, J. C. AND R. D. JAMES. 1975. Aspects of the biology of the Chestnut-sided Shrike- vireo. Wilson Bull. 87:320-334. BOCK, W. J. AND W. DEW. MILLER. 1959. The scansorial foot of the woodpeckers, with comments on the evolution of perching and climbing feet in birds. Am. Mus. Novit. 1931. 76 THE WILSON BULLETIN ß Vol. 93, No. 1, March 1981 CLARK, G. A., JR. 1973. Holding food with the feet in passerines. Bird-Banding 44:91-99. ß 1975. Additional records of passerinc terrestrial gaits. Wilson Bull. 87:384-389. ß 1977. Foot-scutes in North American oscines. Bird-Banding 48:301-308. DORST, J. 1960. A propos des affinities systematiques de deux oiseaux malagaches: Tylas eduardi et Hypositta corallirostris. L, Oiseau 30:259-269. HALL, a. P. AND R. E. MOREAU. 1970. An atlas of speciation in African passerine birds. Br. Mus. (Nat. Hist.), London, England.' HARRISON, C. J. O. 1967. The apparent affinities of lfrita. Bull. Br. Orn. Club 87:97- 100. ß 1969. The affinities of the blue wren genus Malurus and related genera: with special reference to the grass-wren genus Amytornis. Emu 69:1-8. LEISLER, a. 1972. Artmerkmale am Fuss adulter Teich- und Sumpfrohrshnger (Acroceph- alus scirpaceus, A. palustris) und ihre Funktion. J. F. Orn. 113:366-373. MORONY, J. J., JR., W. J. BOCK AND J. FARRAND, JR. 1975. Reference list of the birds of the world. Dept. Ornithology, Am. Mus. Nat. Hist., New York, New York. PARKES, K. C. 1978. Still another parulid intergeneric hybrid (Mniotilta x Dendroica) and its taxonomic and evolutionary implications. Auk 95:682-690. RAIKOW, R. J. 1978. Appendicular myology and relationships of the New World nine-pri- marled oscines (Aves: Passeriformes). Bull. Carnegie Mus. Nat. Hist. 7:1-43. RAND, A. L. AND M. A. TRAYLOR, JR. 1953. The systematic position of the genera Ram- phocaenus and Microbates. Auk 70:334-337. RIDGWAY, R. 1901-07. The birds of North and Middle America. U.S. Natl. Mus. Bull. 50, Pts. I-IV. RGGEBERG, T. 1960. Zur funktionellen Anatomie der hinteren Extremitht einiger mitteleu- rophischer Singvogelarten. Z. Wiss. Zool. 164:1-106. WILLIS, E. O. 1969. On the behavior of five species of Rhegmatorhina, ant-following ant- birds of the Amazon Basin. Wilson Bull. 81:363-395. ß 1972. The behavior of Spotted Antbirds. Ornithol. Monogr. No. 10. ZusI, R. L. 1978. Notes on song and feeding behaviour of Orthonyx spaldingii. Emu 78:156- 157. BIOLOGICAL SCIENCES GROUP, UNIV. CONNECTICUT, STORRS CONNECTI- CUT 06268. ACCEPTED 25 APR. 1980. SEVENTH INTERNATIONAL CONFERENCE ON BIRD CENSUS WORK AND FIFTH MEETING OF THE EUROPEAN ATLAS COMMITTEE An international conference for those involved or interested in bird census and/or atlas work will convene 8-12 September 1981 at the Universidad de Le6n, Spain. For further information write: Prof. Francisco Purroy, Departamento de Zoologia, Facultad de Biologia, Le6n, Spain.