Tracheal Elongation in Birds-Of-Paradise

Condor, 80:423-430 0 The Cooper Ornithological Society 1078

TRACHEAL ELONGATION IN BIRDS-OF-PARADISE

MARY H. CLENCH

Although elongated tracheae are relatively In 1882 Forbes reported on six more speci- well known in certain groups of non-passerines mens of P. keraudrenii: two males with tightly such as swans, cranes, cracids, and guineafowl coiled tracheae; one female with a single (e.g., Portmann 1950), study of their occurrence curved loop (like Pavesis’ specimens) ; another in passerines has been neglected. In this paper female in which the trachea apparently was I shall review what is known about elongated coiled but less so than in males (Forbess’ tracheae in birds-of-paradise (mostly from the wording is ambiguous) ; a third female with an 19th century literature), and report on addi- unelongated trachea (an adult with a mature tional specimens that extend our understand- egg in the oviduct); and a male, “probably ing of the morphology of the phenomenon. In young,” with an unelongated trachea. the hope of stimulating further study, I shall Since Forbes, several additional specimens also offer some speculations on the function have been reported (e.g., Riippell 1933)) but of what is a highly interesting, perhaps even with few details. Recently I examined the spectacular, anatomical specialization. following specimens in alcohol at the Ameri- can Museum of Natural History: three whole MORPHOLOGY adult males (Fig. l), one fledgling, and four The only passerine genera known to exhibit unnumbered “carcasses” from study skins col- elongated tracheae are Phonygammus and lected by E. T. Gilliard (Fig. 2). In addition, Manucodia, closely related birds-of-paradise three study skins in the AMNH collection (Paradisaeidae) . In both, the elongation is have dried tracheae tied to them, and one has superficial, with the looped trachea lying on a label notation describing the tracheal con- the ventral body surface between the skin and figuration. the body musculature, and held in place by Although no one has had a long series in connective tissue. which to study the precise effects of age and sex on tracheal length, the present limited in- PHONYGAMMUS formation (approximately 30 specimens) will support the hypothesis that young birds of An elongated trachea in passerines was first both sexes of Phonygammus have simple, recorded in 1826 in the original description of straight (unelongated) tracheae, and that as the Trumpet Bird (Phonygammus keraudrenii) the birds age, males develop the extremely by Lesson (cited in Pavesi 1874). Lessons’ coiled condition whereas females develop figure and Pavesis’ later studies (1874, 1876) fewer coils-and probably develop them more on nine more specimens show that in this slowly. As a practical matter, the tracheal de- monotypic genus the trachea typically ex- velopment of this species could be studied tends down the neck and then coils in a loop easily in captive birds. The skin is thin and, to the right, regularly or irregularly, over the as the tracheal coils lie primarily under the pectoral muscles. In some specimens the tra- ventral apterium, they are readily visible and chea is so long that the coils extend down over can be measured without doing injury to the the abdominal muscles as well, to cover vir- bird. tually the entire ventral surface of the body. In most individuals the coils form a flat spiral, MANUCODIA but occasionally one coil may lie on top of another. Very long and tightly coiled tracheae In this genus the elongated trachea forms a seem to be found only in adult males (seven single loop that extends a variable distance specimens). In females (two specimens) the down the ventral body surface, but does not looped trachea is shorter, and merely bends coil. The tracheal lengths of even fewer speci- laterally across the pectoral muscles, without mens of this genus have been recorded, but it forming a coil. Pavesi (1876) also noted that appears that length may differ in the four in females the trachea lies in a groove in the species of Manucodia. According to both pectoral muscles, presumably to facilitate Forbes and DAlberti’ (Forbes 1882), the brooding. His specimens may have shown Glossy-mantled Manucode (M. ater) has only this condition through an artifact of preserva- a short loop in the male. The Green-breasted tion: no later authors have mentioned a Manucode (M. chalybatus) was reported by groove. Pavesi and by Salvadori (Pavesi 1876) to be 14231 424 MARY H. CLENCH

FIGURE 1. Phonygunmus keraudrenii (probably jam&i, cf. Crandall 1935); AMNH Ale. Coll. nos. 17, 19, 18, all ex N.Y. Zool. Park. AMNH 18 (C, C)’ was at least 13 years old when it died, its trachea is over- lapped in two places, and is approximately 750 mm long.

similar to the Jobi Manucode (M. jobiensis): forms a short loop of approximately 34 mm. It both have single-looped tracheae that extend agrees well with the specimen of ater figured about % the length of the thorax. Salvadori by Forbes (1882). The carcass of an adult added that in chalybatus the elongation OC- male M. chalybatus collected by Gilliard has curs only in males, that older males have longer a much longer trachea, with the loop extending tracheae than do young males, and that in just beyond the posterior border of the sternum very young males the tracheae are not elon- (Fig. 3). Glli iar d remarked in his field notes gated. In the Curl-crested Manucode (M. that when this specimen was prepared “It had comrii) the trachea is even longer. Beddard the trachea so greatly enlarged that the native (1891) fgi ure d a sp ecimen in which it extends skinner took fright, thinking it was a giant the full length of the ventral body surface be- worm.” A description of this trachea appears fore doubling back on itself to ascend to the in Gilliard and LeCroy (1967). A sketch on furcular region. the skin label of another male chalybatus To this information I can add notes on sev- ( AMNH 427595) shows a similar tracheal con- eral specimens at the AMNH and on one from figuration to that of Gilliards’ specimen-a the British Museum (Natural History). None simple loop extending just posterior to the of the few female alcoholic specimens of any sternal border. species of Manucodia I have examined has Five additional specimens of M. comrii an elongated trachea, and none of the female demonstrate the much greater elongation of study skins in the AMNH have tracheal no- the trachea in this species. One AMNH tations on the label. The single additional carcass has a relatively short trachea that alcoholic specimen of an adult male M. ater curves to the right at the posterior border of consists solely of an excised trachea (AMNH, the sternum (Fig. 4a). Another (Fig. 4b) is unnumbered). This trachea is approximately longer, with the curve well down over the 75 mm long from the posterior end of the abdomen and the end of the loop also extend- glottis to the beginning of the bronchi, and ing up as far as the furculum. In both a third TRACHEAL ELONGATION IN BIRDS-OF-PARADISE 425

A C

FIGURE 2. Phonygammus kerauhenii adelberti. This series suggests the manner in which an elongating trachea begins to coil. Carcasses from AMNH Ale. Coil., all males with enlarged testes; ~011. E. T. Gilliard. carcass from the AMNH and a whole speci- only available specimen with a truly long men from the BM (1905.9.18.9) (Fig. 4~) trachea) may then return to the ventral sur- the trachea descends on the left side of the face, forming folds parallel to the original body, then curves slightly to the right of the course of the trachea. cloaca and along the right side of the tail musculature. It then makes a loop back to the DEVELOPMENT ventral side of the body and extends an- For the mechanism by which these tracheae teriad approximately % the length of the right achieve their remarkable length, we have no side of the body before making a wide loop information from studies of living birds. In- to double back to the cloaca. From the cloaca stead, we can only try to reconstruct the pro- it extends anteriad, packed closely between cess from the limited series of dead specimens the ascending and descending loop on the right that is available. side of the body and the original descending All known young individuals of both genera loop on the left, and enters the body cavity have simple, straight tracheae. The stub- on the right side. Its visible length is 627 mm. tailed fledgling Phonygammus ( AMNH Ale. In another whole adult male of comrii Coll. no. 20) has such an unelongated trachea, (AMNH Ale. Coll. no. 12, Fig. 5) the tracheal approximately 47 mm long, with some 73 configuration on the ventral surface is almost tracheal rings visible without dissection. The identical to that figured by Beddard (1891). outside diameter of this young trachea varies The trachea does not, however, turn to double from 4 to 5 mm, with the rings narrow (0.5 back at the posterior end of the body. It to 1 mm wide) and strongly compressed and continues to extend around the body, passing packed together. The internal lumen diameter to the right of the cloaca and up along the of the trachea is 2.5 to 3 mm. In contrast, an right side of the synsacrum for approximately adult Phonygammus (AMNH Ale. Coll. no. 50 mm. At the level of the right acetabulum 19, Fig. lb) has a trachea that is 828 mm long, it loops tightly back on itself, returns to the and contains approximately 150 rings. Most ventral body surface and then follows the of this birds’ rings are 5 to 7 mm wide, but original descending loop to enter the body some are as much as 22 mm, primarily where cavity on the right side. The total visible the trachea is acutely curved. As with all the length is 511 mm. spirit specimens I examined, the trachea is Thus, in Manucodiu comrii, when a growing hard, but the degree to which this hardness is trachea reaches the end of the body, further a result of preservation is unknown-the tra- growth does not result in a ventral spiral as in cheae of some specimens are slightly more Phonygammus. The extension may lie in part elastic than others. The preserved coil is a along the dorsolateral plane, and (as in the surprisingly rigid structure (Pavesi called it 426 MARY H. CLENCH

which I have been assured by several people fluent in Italian is the equivalent of the Eng- lish inch, and from having to resort to count- ing the number of rings visible in Pavesis’ apparently meticulously drawn illustrations, as he gave few ring counts in his text. ) None of these measurements has been from an in- dividual with an extremely long trachea, but what is more significant is the same lumen diameter reported in a 500 mm-long trachea: 2.5 to 3 mm (Riippell 1933). Unfortunately most of the other measurements available in the literature seem to have been published for their astonishment value (e.g., Ogilvie- Grant 1904)-perhaps understandable in view of the incredible lengths involved. For ex- ample, the 828 mm trachea of the adult Phony- gammus noted above is almost as long as the normal trachea of an Ostrich (Struthio cam&q 910 mm; Hinds and Calder 1971), while Phonygammus has approximately the body size of a Common Flicker (Colaptes auratus, which has a tracheal length of about 38 mm).

DISCUSSION In 1971 Hinds and Calder published a study of the tracheal dead air space in the respiration of birds. They were particularly con- FIGURE 3. Manucodiu chalybatus. Male, cf. Gil- cerned with the increased length of the avian liard and LeCroy 1967. trachea, compared to that of mammals of equivalent body weight, and the consequent a “shield”), with little connective tissue be- increase in mechanical resistance of air flow tween the closely apposed segments. I was within the longer cylinder in long-necked ani- able to make approximate measurements on mals. They showed that, in general, birds two of the Phonygammus carcasses with rela- compensate for the added resistance by having tively short and uncoiled tracheae: 90 rings a greater tracheal diameter (lumen width) total 315 mm in length (Fig. 2c) and 100 and hence greater air volume in that trachea rings total 300 mm (Fig. 2d). The outside (4% times that of mammals of equal body tracheal diameter of all seven AMNH adults mass), in turn partially compensated for by in alcohol averages about 3.5 mm, but the a lower respiratory rate. Non-passerine birds lumen is 2.5 to 3 mm: the same as in the with elongated tracheae, however, have much juvenile. Hence it may be reasonable to as- greater tracheal volumes for their body mass sume that as a bird ages, the trachea lengthens than do those with straight tracheae: Whis- both by differential longitudinal growth tling Swan (Olor columbianus) 2.4 times (widening) of the rings already present, and greater: Trumpeter Swan (0. buccinutor) by the formation of additional rings. During 3.8~; and Sandhill Crane (Grus canadensis) this process, although the walls of the tracheal 3.2~ in the male, 2.4~ in the female-com- rings become thinner as the ring widens, the pensating for the increased tracheal length diameter of the lumen seems to remain con- by an increased width of the lumen. stant at 2.5 to 3 mm. Hinds and Calder did not measure a bird-of- The few other Phonygammus specimens paradise, although they were aware of the con- whose tracheal lengths have been reported dition: “The trumpter manucode and the have all been adult males: three specimens whooping crane have also inherited a respira- have tracheae ranging from about 450 to 550 tory inexpedience of similar or greater se- mm long, and containing from about 100 to verity [than the Trumpter Swan]” (1971:438). 120 rings. (My uncertainty here results both Hinds and Calder were using (usually) fresh from Pavesis’ measurements being in pollici, specimens, measuring the length of the tra- TRACHEAL ELONGATION IN BIRDS-OF-PARADISE 427

FIGURE 4. Manucodia comrii. A, B: carcasses from AMNH Ale. Coll.; C: BM 1905.9.18.9. chea from the glottis to the syrinx, and com- of 12 specimens of P. keruudrenii (males only, paring tracheal data with body weights. For three subspecies pooled) is 166.2 g, range 130- the paradisaeids, I have seen only preserved 190 g. Unfortunately I have been unable to specimens of unknown fresh weights, and a locate any weights for M. comrii. M. ater few weights of other individuals recorded on males average 219.4 g (n = 7)) M. chalybatus, the labels of study skins. To avoid damage to 236 g (n = 6), and a single M. jobiensis, 232 rare, perhaps irreplaceable specimens, my g; but M. comrii is considerably larger than measurements of tracheal length have been the other three species in this genus (male limited to the visible portion of the trachea: comrii wing measurements range from 219 from the glottis in carcasses or from near the to 248 mm, compared to 159 to 205 for the glottis in whole but skinned alcoholics, to the others [Gilliard 19691). Hinds and Calder furcular region where the trachea enters the reported the tracheal length of a preserved body cavity. Hence calculations exactly com- Whistling Swan as 876 mm (vs. a visible parable to those of Hinds and Calder cannot length of 828 in the Phonygammus), yet its be made. I can, however, suggest that tra- tracheal diameter was 10.8 mm (vs. 2.5 to 3 cheal (lumen) diameters of 2.5 to 3 mm in mm in Phonygammus). The same measure- preserved Phonygammus keraudrenii and 4 to ments in preserved Trumpeter Swans were 6 mm in Manucodia comrii do not sufficiently 1210-1350/12.2-12.3 and in preserved Sand- compensate for the elongated tracheae in these hill Cranes, 605691/8.1-9.0. As Hinds and birds, compared to other passerines. I make Calder showed, preservation reduces tracheal this suggestion from the following data: single dimensions slightly, hence these swan and fresh specimens of Black-billed Magpie (Pica crane figures as well as those from the birds-of- pica) and (presumably) Great-tailed Grackle paradise above represent minima for living (“Cassidix mexicanus”) measured by Hinds birds. It is therefore clear that Phonygammus and Calder both had 3 mm lumen diameters. and Manucodia have approximately the same These species have a body size comparable tracheal lumen measurements as “normal” to that of Phonygammus. The weight of the passerines of similar body weight, and much magpie used by Hinds and Calder was 156.2 smaller lumens than non-passerines with g; that of the grackle, 159 g. The mean weight elongated tracheae. 428 MARY H. CLENCH

not subject to the surface tension problems found in mammalian lungs. Hence the energy consumed in moving air across those surfaces is considerably less. In addition, the flow across the exchange surface is unidirectional, and the ability of birds to extract oxygen greatly exceeds that of mammals. Also, with the air sac system occupying as much as 25% of the volume of the body cavity, the potential tidal volume is enormous, especially in comparison with the dead air space in the trachea. Thus the amount of tracheal dead space is of comparatively less importance to birds than to mammals, except during periods of strenuous and prolonged activity when it could become critical. It may be that these birds-of-paradise can “afford” a greatly elongated trachea because they are sedentary (non-migratory) and, being relatively large forest birds, their daily flight distances are probably short and predators can be avoided by short dodging flights around foliage. All of these assump- tions remain to be tested. FIGURE 5. Manucodia comrii. AMNH Ale. Coil. no. 12; ventral and right lateral aspects. VOCALIZATIONS One can also speculate on how this extreme tracheal elongation affects the vocalizations There is also a serious discrepancy in com- of the living birds. As with so many New parative tracheal volumes. According to my Guinea and tropical Australian species, little rough estimates, the fledgling Phonygammus is known of the habits of with a straight trachea had a minimum dead Phonygammus and Manucodiu. Following the conclusions of air volume of 0.33 cc in its trachea; the adult Marion (1977) who studied the elongation of (AMNH Ale. Coil. no. 19) had a min. 5.85 cc. the trachea in maturing male Plain Chacha- The Manucodia (AMNH Ale. Coil. no. 12) lacas had a min. 10.03 cc. For comparison, Hinds (Ortalis vetula), it is reasonable to as- and Calder reported a tracheal volume of sume that the tracheal elongations of birds-of- 0.48-0.49 cc in the magpie and grackle and paradise also serve to lower the pitch of, and 80.25 cc in the Whistling Swan. perhaps amplify, their vocalizations. Concerning Phonygammus, Thorpe (quoted Phonygammus and Manucodia may have in Gilliard 1969: 105) some additional form of compensation-phys- wrote “The males utter iological, behavioral, or morphological-to a very loud and deep guttural note, unlike compensate for the high volume of dead air in that of any other bird I am acquainted with, and it astonished me that a comparatively their tracheae. I have searched for a connec- tion between the ascending and descending small bird could make so much noise.” Ralph Bulmer (field notes on file at the AMNH) re- portions of the tracheae, but have found no shunt mechanism that could act to bypass marked that Phonygammus is “a very noisy, obvious, tree-top feeding bird.” Gilliard (1969) the elongated section, A compensating mech- also characterized the species as having anism, if it exists, might be found through study of living birds. “powerful call notes.” It is also possible that a compensating mech- I have been able to find few descriptions anism does not exist. Hinds and Calder paid of the calls produced by Manucodia. The insufficient attention to the strong differences vocalizations of M. ater include whistles and between avian and mammalian respiratory a deep “chug” call (Rand 1938) and a “long- systems, especially the extensive air sac sys- drawn-out plaintive whistle” either on one tem in birds. Abbot S. Gaunt advises me note (Mackay 1970:52) or tremulous (Anon. (pers. comm.) that birds are much less sus- 1972). M. jobiensis, and possibly also M. ceptible to the problems of dead air space ater, has a “long-drawn moaning cry” (Gil- than are mammals. The expansile portions of hard 1969). M. chalyhatus has a staccato the avian respiratory system are large and are “tuk” call (Diamond 1972) and “a deep short TRACHEAL ELONGATION IN BIRDS-OF-PARADISE 429 whoo‘ ,’ like a pigeon or coucal, given after SUMMARY great swelling of the chest, rather like the The morphology of the elongated tracheae in Trumpet-bird” (Anon. 1972:5). Gilliard also two genera of paradisaeids is described. In remarked that, as M. jobiensis and chalybatus Phonygammus the trachea is unelongated in are sympatric over much of their range, and young birds but as they age, males develop as their plumage is very similar, their isolat- a long spiraled tracheal coil on the ventral ing mechanisms must be “almost entirely be- body surface beneath the skin; females de- havioural (probably chiefly vocal)” ( 1969: velop a shorter loop or coil and probably de- 100). The vocalizations of M. comrii do not velop it more slowly. In Manucodia the tra- seem to have been reported. Information chea lengthens as a loop rather than a coil. It about them would be particularly useful, for has been found only in adult males and is this manucode exhibits the greatest elonga- probably species-specific, with minimum de- tion of the trachea. velopment in M. ater and maximum in M. In a description of captive “Manucode” comrii. The longest trachea recorded in (probably Phonygammus) vocalizations, Seth- Phonygammus is 828 mm; that in Manucodia Smith ( 1923:56) reported: “[the coiled tra- comrii is 627 mm (both minimum measure- chea] doubtless accounts for the peculiar note ments) . of these birds, which, although, not loud, is Speculations are offered on the physiolog- penetrating and audible at a considerable ical implications of this tracheal specialization. distance. Before uttering this the bird raises Both genera have approximately the same tra- its wings and extends the body, the note be- cheal width (lumen diameter) as other pas- ing emitted as the wings are lowered and the serines of similar body weight, and much body contracted, the sound being, as it were, smaller lumens than non-passerines with com- pumped out of the body through the specialized trachea.” Both this description and that parably elongated tracheae. The elongations of Manucodia chahybatus (Anon. 1972) sug- may or may not be a serious physiological gest that a call is produced only after a bird liability. Comments are also made on the pos- has taken an especially deep breath. sible effects of tracheal elongation on these birds ’ vocalizations. TAXONOMIC POSITION ACKNOWLEDGMENTS Phonygammus and Manucodia were long con- sidered generalized and crow-like primitive Specimens for this study were borrowed from the members of the Paradisaeidae (cf. Stonor American Museum of Natural Historv and the British 1938). This was chiefly because they lack Museum (Natural History) through- the courtesy of Walter Bock and Philip Burton, respectively. Mary specialized courtship and breeding behavior; LeCroy kindly made available the field notes of the are not sexually dimorphic; and have rela- late E. T. Gilliard and others at the AMNH, and tively simple black plumage, although some of lent me her expertise in the New Guinea literature. the feathers are structurally modified. In his Felix Wiedenmayer and Arthur Bianculli were of in- valuable help in translating archaic Italian. Kenneth study of the relationships of the birds-of-para- Parkes assisted me in looking through the paradise and bowerbirds, Bock (1963) empha- disaeid studv skins at the AMNH when time was sized that, from skull characters, Manucodia limited, and-gave valuable advice on the manuscript. and Phonygammus are not primitive but are Harry Clench checked my calculations of tracheal as specialized as the ornately-plumed Para- volume and also read the manuscript. Nancy J. Perkins prepared the line drawings. To all I ex- disaea. He also speculated that the plain tend my thanks. plumage and normal courtship and breeding Abbot Gaunt generously shared his experience habits are secondarily primitive, with the with avian respiratory systems. He also offered the similarities to corvids being convergent. From suggestion that Phonygammus and Manucodia are the tracheal evidence (which Bock also men- not really birds, but clever wind-up toys: the sup- posed elongated trachea is actually a power spring tioned) I agree that these two genera are, in- that has been enclosed in a thin membrane to keep deed, highly specialized. Unlike the rest of it waterproof. “Males” are simply the wound-up the family, however, these birds ’ specializa- version, “females” the run-down version. tions seem to lie primarily in their vocal structure and behavior-particularly in the LITERATURE CITED

quality, pitch, and volume of sound. If a ANONY~MOUS. 1972. Third Ornithlon results. New greatly elongated trachea is a physiological Guinea Bird Sot. No. 74:1-5. liability, the vocalizations it produces must BEDDARD, F. E. 1891. Ornithological notes. II. On the convoluted trachea of Manucodia comrii. be sufficiently compensatory to the species Ibis ( 1891) :512-514. to be adaptive: BOCK, W. J. 1963. Relationships between the birds 430 MARY H. CLENCH

of paradise and the bower birds. Condor 65: the Plain Chachalaca in south Texas. Wilson 91-125. Bull. 89:47-56. CRANDALL, L. S. 1935. “The most beautiful birds OGILVIE-GRANT, W. R. 1904. [Meeting report.] in the world,” an account of the birds of para- Bull. Br. Ornithol. Club 14:40-41. dise in public and private collections. Bull. N.Y. PAVESI, P. 1874. Intorno ad una nuova forma di Zool. Sot. 38: 147-160. trachea di Munucodiu. Ann. Mus. Civ. Stor. Nat. DIAILIOND, J. M. 1972. Avifauna of the eastern Genova 6:315-324. highlands of New Guinea. Publ. Nuttall Ornithol. PAVESI, P. 1876. Studi anatomici sopra alcuni uccelli. Club No. 12. Ann. Mus. Civ. Stor. Nat. Genova 9:66-82. FORBES, W. A. 1882. On the convoluted trachea of PORTMANN, A. 1950. Les organes respiratoires. In two species of manucode (Manuco&u at~u and Gras&, P. red.], Traiti: de zoologie 15:257-269. Phonygamn gouldi): with remarks on similar RAND, A. L. 1938. Results of the Archbold Expe- structures in other birds. Proc. Zool. Sot. Lon- ditions. No. 22. On the breeding habits of some don ( 1882) :347-353. birds of paradise in the wild. Am. Mus. Novitates GILLIARU, E. T. 1969. Birds of paradise and bower 993:1-8. birds. Natural History Press, New York. R~~PPELL, W. 1933. Physiologie und akustik der GILLIARD, E. T. AND M. LECROY. 1967. Annotated vogelstimme. J. Omithol. 81:433-542. list of birds of the Adelbert Mountains, New SETH-SWTH, D. 1923. The birds of paradise and Guinea. Bull. Am. Mus. Nat. Hist. 138:51-82. bower birds. Avic. Maa. Ser. 4. 1:41-60. HINDS, D. S. AND W. A. CALDER. 1971. Tracheal STOITOR, C. R. 1938. Some features of the variation dead space in the respiration of birds. Evolu- of the birds of paradise. Proc. Zool. Sot. Lon- tion 25: 429440. don 108:417-481. MACKAY, R. D. 1970. The birds of Port Moresby Carnegie Museum of Natural History, Pittsburgh, and district. Nelson, London. Pennsylvania 15213. Accepted for publication 21 MARION, W. R. 1977. Growth and development of March 1978.