Specializations of Some Carotnoid-Bearing Feathers

SPECIALIZATIONS OF SOME CAROTENOID-BEARING FEATHERS

STORRS L. OLSON

Department of Pathobiology

School of Hveiene_I and Public Health The Johns Hopkins University Baltimore, Maryland 21205

Certain of the Eurylaimidae (broadbills) the barb was composed of distinct light- and Cotingidae (cotingas) bear feathers of a colored cells. There was, however, a small violet or deep maroon color that is uncommon area of unorganized dark material at the in birds. These feathers show a number of larger end of the sections. Traced proximally peculiarities not found in typical feathers through serial sections, this dark area ex- colored by melanins or schemochromes (struc- panded until, in the region where barbules tural colors). I investigated the external and started to appear, it was discerned as the internal morphology of feathers from broad- medulla and was composed of distinct cells bills, cotingas, and other families in order to (fig. 1B). Further proximally, where the determine details of structure, amount of barbules reached their fullest development, variation between species, and the possible the medulla constituted an even larger area taxonomic value of these specialized feathers. (fig. 1C). The cortical cells, which were In a typical feather barb there are rows of quite distinct in distal sections, became dif- barbules present. Cross sections show two fuse and indistinguishable. A thick, wavy layers of cells: a prominent pigmented me- layer surrounding the outside of the barb, dulla surrounded by a cortex that usually in which no cells could be discerned, was appears clearer and more or less homo- particularly noticeable in the distal sections. geneous. Desselberger ( 1930) pointed out This layer, which was present in all the that when heavy deposits of lipochromes are feathers I examined, is what Strong (1952) present, barbs may be flattened, lack barbules, mistakenly alluded to as the cortex in QueruZu and be undifferentiated internally. Brush and purpurata. Strong (1902) earlier referred to Seifried (1968) showed a definite correlation the possible presence of an “epitrichium” (an between this modified structure and carote- inappropriate term for a feather component) noid deposition in the Gouldian Finch surrounding the barb cortex but regarded it (Poephilu gouldiue) and several other species. as purely an optical effect. That such is not the case is vividly demonstrated in Dryocopus METHODS pileatus (see below). Hereafter, for want of Feathers, obtained from museum skins, were selected a better term, I shall refer to this layer as the on the basis of their color and gloss. The gross cuticle. structure of barbs and barbules was examined micro- Guianan Cock-of-the-Rock. Rupicolu rupi- scopically. Feathers were embedded in nitrocellulose cokz. In the orange crown of males there is (Parlodion), following generally the methods out- lined in Galigher and Kozloff ( 1964). Paraffin was a subterminal band of deep brownish maroon not used as feathers embedded in it tended to shatter that illustrates particularly well the special- when cut. The entire feather was sectioned trans- izations associated with heavy depositions of versely on a rotary microtome from the distal carotenoids. Under magnification (fig. 2A), towards the proximal end. Sections were made at the orange portions of the barbs both proximal varying thicknesses,usually about 20 ~1. Unless other- wise stated, the descriptions were taken from the and distal to the maroon band were not ex- modified distal parts of the barbs. panded and bore barbules which were normal in size and number. However, on the maroon RESULTS band the barbs were flattened and the bar- COTINGIDAE bules, although still present, became vestigial. Sections through the maroon portion showed Red-ruffed Fruitcrow. Pyroderus scutatus. it to consist almost entirely of distinct cortical The throat and upper breast feathers are orange, tipped with a deeper scarlet-orange, cells, with only a few dispersed dark granules and they present a singular crinkled appear- of melanin indicating the location of the re- ance. The distal tips of the barbs lacked mains of the medulla (fig. 2B). In the orange barbules, were conspicously flattened, and it portions (either proximal or distal), the was here that the color was most intense. In medullar area increased in size and orga- distal sections (fig. 1A) the greater part of nization and the cortical cells became less

The Condor, 72:424-430, 1970 14241 SPECIALIZATIONS OF SOME CAROTENOID-BEARING FEATHERS 425

6 C ..

FIGURE 1. Cross sections from a barb of a throat feather of Pyroderus scututus. A. Section from modified distal portion of barb. B. Section taken between A and C approximately where barb&s commence. C. Proximal section taken through area of well developed barbules. All sections are shown ventral side uppermost. m = medulla; c = cortex; b = barbule.

distinct (fig. 2C), and in the region of normal ment of specialized carotenoid-containing barbule development, the medulla assumed feathers of any group of birds. Except for the its full stature and the outlines of the cortical remiges and rectrices, the males are adorned cells were lost (fig. 2D). entirely with glossy feathers ranging from Purple-throated Fruitcrow. Querula pur- deep maroon to blackish-purple. In the pumta. The maroon gorget feathers of the Pompadour Cotinga (X. punicea) the barbs male bear long, flattened, barbuleless barbs. were much elongated and lacked barbules There have been conflicting reports of the for at least two-thirds of their length. Distally, internal structure of these barbs. Strong the rachis itself was pigmented and became (1952) described cross sections as having a indistinguishable from the barbs, as at this thin outer layer surrounding a “homogeneous, point the branching became dichotomous, pigmented, central core . . . in which no trace rather than pinnate as is usual in most of medullary cells could be found,” whereas feathers. As in Querulu, there was no medulla Mattem (1956) reported cells in the pig- and the cortex revealed no definite cellular mented portions and large air-filled medullary structure. cells in proximal sections. She believed Cotinga. Males of this genus are bright Strongs’ observations resulted from fading blue, with patches of deep violet below. The of the pigments in the embedding medium he violet throat feathers from the Lovely Cotinga used, causing the cell outlines to become in- (C. amabilk) were glossy with flattened distinct. My observations confirm Strongs’ barbuleless barbs. Sections lacked a medulla; I could not distinguish cells in barbs of distinct cortical cells were present (fig. 3A). Querula (or Xipholena), nor did I find air- Some of the violet feathers, usually those of filled medullary cells. In a very few sections the belly, did not appear glossy and had there were scattered small, dark remnants of normal barbule development. In sections of medulla, such as illustrated in figure 2B for barbs from them (fig. 3B), a small medullar Rupicolu. The outer layer mentioned by spot was seen. This consisted of a dark area Strong represents the cuticle, the “core” being within a lighter area and contained no dis- the cortex. tinct cells. The cortical cells were apparent. Xiphokna. The three species of Xipholena Phoenicircus. Red crown feathers from exhibit perhaps the most extensive develop- males of the Guianan Red-Cotinga (P. carni- 426 STORRS L. OLSON

/ I /

/ // / I / 1 ’ / / ’ 1 ’ / 1 ’

m m

C .6. FIGURE 2. A. Distal portion of a single barb from the crown of a male Rupicola rupicola; shaded portion is maroon, unshaded portions, orange. B-D. Cross sections of barb; dotted lines show approximate levels of the sections. All sections shown ventral side uppermost. m = medulla; c = cortex; b = barbule.

fex) and the Black-necked Red-Cotinga (P. Red-banded Fruiteater. Pipreolu whitleyi. nigricollk) were examined. The distal barbs Feathers from the reddish-orange breast band lacked barbules; proximal barbs bore barbules of a male revealed flattened distal barbs with- at the base. No traces of a medulla were out barbules. There was a fairly large medul- found in P. nigricollis, while a few proximal lar spot at the larger end of the sections and sections of P. carnifex did show a dark central the cortical cells were in evidence. medullar vestige. Cortical cells were evident Purple-throated Cotinga. Porphyroluema in both species. porphyroluemu. Glossy, deep violet feathers

FIGURE 3. A. Cross section of barb from glossy throat feather of Cotingu amabilis. B. Cross section of barb from non-glossy breast feather of Cotinga umubilis. C. Cross section of barb from cheek feather of Cymbirhynchus mucrorhynchus. D. Distal end of barb of cheek feather of Cymbirhynchus mucrorhynchus. E. Cross section of barb from crown of Pipru uureokz. F. Cross section of barb from red feather of Phlogothruupis sunguinolentu. G. Cross section of barb from crest of Dryocopus pileutus. None to scale. All sections shown ventral side uppermost. m = medulla; c = cortex; b = barbule; cu = cuticle. SPECIALIZATIONS OF SOME CAROTENOID-BEARING FEATHERS 427

C D

E F G 428‘ STORRS L. OLSON from the throat of a male bore flattened barbs THRAUPIDAE lacking barbules distally. Sections appeared Crimson-collared Tanager. Phlogothraupis more granular than in other species examined. sanguinolenta. The glossy red feathers of this A medullar spot could be found in only a species presented somewhat flattened barbs few sections, but cortical cells were distinct with at least rudimentary barbules through- in all. out their length. Cross sections revealed a PIPRIDAE definite dark medullar spot. Outlines of cortical cells could be distinguished only in the Crimson-hooded Manakin. Pipra aureola. pointed end of the sections (fig. 3F). Crimson crown feathers from a male bore barbs which were flattened distally and PICIDAE lacked barbules. Cross sections were elongate Pileated Woodpecker. Dyocopus pileatus. and pointed at one end. No medullar spot The glossy red crest feathers bear very long, was perceived; cortical cells were obvious barbuleless barbs, In section these presented (fig. 3E). a most singular appearance. There was an Helmeted Manakin. Antilophia galeata. area of distinct cortical cells which in some The deep crimson feathers from the peculiar of the sections contained a small, dark medul- crown of males also displayed flattened, lar spot. Surrounding this cortex was a barbuleless barbs. In only a few proximal greatly enlarged and thickened cuticle (fig. sections was there a vestigial medulla. The 3G) which appeared to be tinted purplish in cortex was composed of distinct cells. reflected light. In this translucent cuticle, no structures were seen that would indicate any EURYLAIMIDAE cellular composition. Black and Red Broadbill. Cymbirhynchus macrorhynchus. On the rump, lower throat, DISCUSSION and abdomen are glossy feathers of a deep Viilker (1952, 1954) and Brush (1969) maroon color. The distal portions of the barbs found that various carotenoids are responsible lacked barbules and in the cheek region, for the red and violet colors in cotingas. where these feathers are very plush and Desselberger ( 1930)) Frank ( 1939)) Brush glossy, the barbuleless parts are markedly ex- and Seifried ( 1968), and others have cor- panded and flattened (fig. 3D) There was related heavy carotenoid depositions with a a distinct, fairly large medullar area (fig. 3C) highly modified feather structure in which although this varied in size and was absent in barbs are markedly flattened and lack bar- some sections. Cortical cells were apparent bules. This condition was found in most of in all sections. the glossy, highly pigmented feathers I ex- Wattled Broadbill. Eurylaimus steerei. The amined. The loss or reduction of barbules crown and upper rump feathers are a deep and the expansion and flattening of the barb purplish-maroon, almost brownish in appear- are specializations that presumably accomp- ance. The barbs of the rump feathers were lish a more effective display of the pigments not as flattened as those of the crown. In than could be gained with typical feather both, barbules were absent, a small medullar structures. Frank (1939) noted that lipo- spot was present, and the cortical cells were chromes are primarily deposited in barbs as very distinct. opposed to melanins which are commonly Banded Broadbill. Eurylaimus jauanicus. found in barbules. In some barbs of the dark brownish-maroon There are corresponding internal modifi- crown feathers barbules were lacking, while cations of the barb where these external in others they extended to the tip of the specializations are found. Serial sections of barb. The sections were more rounded in barbs from Pyroderus scutatus and Rupicoh shape than for other barbs seen. No trace of rupicola showed that the medulla, which a medulla could be found, though cortical plays an important role in structural color- cells were in evidence. ation, becomes vestigial or is lost altogether in the carotenoid-bearing portions of these barbs. TYRANNIDAE In the other species investigated, the medulla Vermilion Flycatcher. Pyrocephalus rubi- was either absent or appeared sporadically as nw. Barbs of the glossy vermilion feathers a small vestigial structure. The carotenoid from the crown of a male lacked barbules pigments seem to be deposited exclusively in distally. No vestige of a medulla was found the cells of the cortex. In the modified tips in any of the sections. of the secondaries in the Cedar Waxwing SPECIALIZATIONS OF SOME CAROTENOID-BEARING FEATHERS 429

(Bombycilla cedrorum) , the carotenoid pig- cells in the cuticle of any of my specimens ments are also deposited in the cortex of the (even DT~JOCO~US), Auber and Appleyard rachis (Brush and Allen 1963). Strong (1902: (1951) have described and pictured the sur- 161) stated that in the fully comified feather face layer of the cortex as composed of flat- barb “there is little or no evidence of” the tened cells resembling the cuticle of mam- “former cellular nature” of the cortex. This, malian hair. They also found (1955) a however, seems to apply mainly to schemo- cellular cuticle in the honeycreepers Chloro- chromatic or melanin-pigmented feathers, phanes and Iridophunes. The presence of a whereas in carotenoid-containing feathers the well-developed cuticle in carotenoid-laden cortical cells are usually quite evident. In feathers probably explains their high gloss. some forms (e.g., Querulu, Xipholena) the Many carotenoid-bearing feathers have a cellular composition of the cortex seems to propensity to change color with heat or pres- be abrogated. Serial sections of Pyroderus sure. This was first reported by Quelch and Rupicola barbs show that the cortical (1896) in Cotinga and Xiphobna. Color cells degenerate progressively as the medulla change was reported by Frank (1939) for increases in size. This may indicate that the Picus viridus, by Strong (1952) for Querula, presence of the medulla in some way by Mattem and Volker (1955) and Mattem suppresses the formation of cortical cells or (1956) for a number of other species. This causes them to degenerate. Conversely, the change is not marked in reddish or orange presence of great quantities of carotenoids feathers, which only become lighter orange. may inhibit formation of the medulla. However, deep maroon or violet feathers such Concerning the development of these types as found in Xipholena, Querula, Cotinga, of feathers, Voitkevitch (1966:29) noted that, Porphyrolaemu, and Cymbirhynchus, when Where‘ great quantities of lipochromes ac- heated near a flame, change dramatically to cumulate during structure differentiation, the a light yellowish-orange. The nature of violet cell material of the first order barbs [=barbs] coloration in cotingas has been subject to dis- is stimulated to a more intensive growth than pute. GSmitz and Rensch (1924) attributed that of the second order barbs [=barbules]. this color to a pigment which they called Consequently, the parts of the first order “c0tingin.” However, the extracted pigment barbs colored by lipochromes (or with mel- from violet feathers of Xipholena punicea is anins) are relatively more bulky.” However, brick red and consists of a complex of several contrary to what is stated, this does not apply carotenoids (Brush 1969). The pigments to melanins. themselves are not violet and there is no The lack of barbules, at least in some spe- typical schemochromatic medullary structure cies, is a factor of wear. Dwight (1900) present in violet feathers to account for the showed that in Carpodacus purpureus and blue portion of the color. Mattem (1956) Loxiu curvirostra the feathers of the bright and Schmidt ( 1956) reported that violet color definitive male plumage bear barbules that is a result of the alignment of small granules are worn away to expose the flattened barb of pigment. Heat destroys the granules and during the breeding season. Desselberger pressure causes them to realign, producing (1930) described weak deciduous barbules the change in color. Brush (1969) observed that are soon lost from the red mask of the that the red extract of feathers of Xiphoknu European Goldfinch ( Carduelis carduelis). punicea changed to yellow when treated with While wear is responsible for the barbuleless sodium borohydrate and that the keto-carote- condition in some species, in others the barbs noid curve observed prior to treatment apparently develop without barbules. For in- changed to one typical of hydroxy-carotenoids. stance, the barbs of the gorget feathers of This may explain why heated feathers turn Querulu show no trace of barbules, even yellowish rather than red. I can find nothing while still in their sheaths. in the literature or my own observations to Surrounding the cortex is a clear, cuticular support Brushs’ (1969:431) statement that layer of varying thickness. This cuticle is not “the violet coloration of certain species (e.g., an effect of optics as suspected by Strong Cotinga cotinga) is considered to be produced (1902). In the crest feathers of Dryocopus structurally and is not a true pigment.” The pileatus the cuticle constitutes a major por- violet feathers from Cotinga are similar to tion of the barb. A cuticle was also observed those of Xipholenu in appearance and physical in the specialized waxy portions of the second- behavior; i.e., they change color when heated aries of the Cedar Waxwing (Brush and or subjected to pressure. I should emphasize Allen 1963). Although I could not distinguish that not all feathers that .contain carotenoids 430 STORRS L. OLSON are necessarily specialized in the manner of AUBER,L., AND H. M. APPLEYARD.1955. The struc- those discussed in this paper. Many appear ture of the feathers in Chlorophanes and Iti- dophanes (Coerebidae). Ibis 97:252-258. to have no external modifications. The special- BRUSH, A. H. 1989. On the nature of “‘cotingin.” izations described seem to be associated with Condor 71:431433. very heavy depositions of carotenoids in dis- BRUSH,A. H., AND K. ALLEN. 1963. Astaxanthin in tinctive patterns of display, although again the Cedar Waxwing. Science 142:4748. there are exceptions. The presence of these BRUSH,A. H., AND H. SEIFRIED. 1988. Pigmentation and feather structure in genetic variants of the specialized feathers in a number of unrelated Gouldian Finch Poephilu goulcliae. Auk 85: avian taxa indicates that this feature carries 416430. little or no taxonomic significance. DESSELBERGER,H. 1930. Ueber das Lipochrome der Vogelfeder. J. Omithol. 78:32%376. SUMMARY DWIGHT, J. 1900. The sequence of plumages and moults of the passerine birds of New ‘York. Ann. Certain feathers with heavy depositions of New York Acad. Sci. 8( 1):73-360. carotenoids exhibit distinctive modifications. FRANK, F. 1939. Die F&bung der Vogelfeder durch The barbs are flattened and usually lack Pigment and Struktur. J. Omithol. 87:426-523. barbules. The internal structure of the barbs GALIGHER,A. E., AND E. N. KOZLOFF. 1964. Essen- tials of practical microtechnique. Lea and Febi- differs from schemochromatic feathers in that ger, Philadelphia. the medulla is reduced or absent, there are GBRNITZ, K., AND B. RENSCH. 1924. Ueber die distinct cortical cells which contain the pig- violette F&bung der Vogelfedem. J. Omithol. ments, and a thickened exterior cuticle is 72:113-118. MATTERN, I. 1956. Zur Histologie und Histochemie present, giving the barbs a glossy appearance. der lipochromatischen Federn einiger Cotingiden These modifications are believed to increase (Schmuckvogel). Z. Zellforsch. 45:96-136. the effectiveness of carotenoid display. They MA~EXIN, I., AND 0. Vij~xzn. 1955. Die schwartz- are found in many groups of birds and do not roten und violetten liposchromatischen Feder- indicate phylogenetic affinity. farben der Cotingiden. Naturwissenschaften 42: 612-613. ACKNOWLEDGMENTS QUELCH,J. J. 1896. Abstract of address on the birds of British Guiana. In Papers presented to the Portions of this paper were extracted from a master’s World’s Congress on Ornithology. Charles H. thesis filed at Florida State University. In that form Serge1 Company, Chicago. it was criticized bv R. K. Godfrey. R. B. Short, H. SCHM~T, W. J. 1956. Polarizationsoptik und Far- M. Stevenson, andk. W. Yerger. -To Dr. Short I am berscheinungen der lipochromfuhrenden Fede- especially indebted for the use of his laboratory and raste von Xivholena lamellipennis. Z. Zellforsch. equipment and for his advice on microtechnique. To 45: 152-175.’ Alan H. Brush, who generously read and commented STRONG,R. M. 1902. The development of color in upon the manuscript in its many stages of develop- the definitive feather. Bull. Mus. Comp. Zool. ment and who provided much useful information, I 40: 147-186. extend my appreciation. Peter Stettenheim read the STRONG,R. M. 1952. A peculiar pigmentation. Auk manuscript and supplied several references and many 69: 199-200. helpful suggestions. For the use of specimens and VOITKEVITCH,A. A. 1966. The feathers and plumage facilities, I would like to thank Richard L. Zusi and of birds. October House, New York. the U. S. National Museum. Dr. and Mrs. Alexander VijLKER, 0. 1952. Die Lipochrome in den Fedem Wetmore kindly provided accommodations in Wash- der Cotingiden. J. Omithol. 93:122-129. ington during my research there in 1967. V~LKER, 0. 1954. Rhodoxanthin und Zeaxanthin als Pigmente in den Fedem von Cotingiden. Natur- LITERATURE CITED wissenschaften 38:565. AIJBER, L., AND H. M. APPLEYARD. 1951. Surface cells of feather barbs. Nature 168:736. Accepted for publication 16 April 1970.