Jap. J. Ornithol. 40: 85-91, 1992

Tongue of Two of from the Philippines, with Notes on Feeding Habits of (Dicaeidae)

Hiroyuki MORIOKA

Department of Zoology, National Science Museum, Hyakunin-cho 3-23-1, Shinjuku-ku, Tokyo 169

フ ィ リ ピ ン産Prionochilus属 ハ ナ ド リ類 二 種 の 舌 と ハ ナ ド リ科 の 採 食 習 性

森 岡 弘 之 国立科学博物館 動物研究部

The Prionochilus consists of six species of Malaysian and Philippine flowerpeckers with a relatively large tenth primary and a short and stout bill. The six species are very similar to each other and form a natural group. The genus is presumably very closely related to (MAYR & AMADON1947, SALOMONSEN1960). The tongue of flowerpeckers was summarized by GARDNER (1925). He wrote that "the Dicaeidae have small tongues that are flat posteriorly but at about the middle become abruptly narrower and begin to curl into a semitube which is deeply cleft at the tip, the margins of which are smooth, forming two slender semitubular tips. This is found in Dicaeum cruentatum, D. sanguinolentum, D. flammeum [=D. trochileum], and D. celebicum. In Acmonorhynchus [ =Dicaeum ] aureolimbatus the same holds true except that the edges of each tube show a slight notching, with an attempt at the production of four tips, while in Prionochilus these notches have deepened to actual splitting with the formation of four semitubular fringeless projections." GARDNER's statements have been cited and provided a basis for certain statements of subsequent authors who discussed the relationships and evolution of flowerpeckers (Dicaeidae), such as MAYR & AMADON(1947) and RAND (1961, 1967). Thus MAYR & AMADON(1947) stated that in Dicaeum and Anaimos [=Prionochilus] the tongues are, in varying degrees, tubular as in other in which nectar is an important item of the diet and they envisaged the evolution of the tongue from a generalized type with frugivorous habits such as in Melanocharis and Paramythia to a more specialized one adapted for nectar feeding (Dicaeum and Prionochilus). I have examined specimens of Prionochilus platen and P. olivaceus and found that these two species possess an unspecialized, rather flat and fleshy tongue similar to those of Melanocharis and Paramythia. I therefore redescribe it below, with some comments on the feeding habits of flowerpeckers. GARDNER (1925) figured a number of tongues for diverse taxa of birds, but he did not illustrate any for the Dicaeidae, giving only descriptions. Nor did he mention the species of Prionochilus on which his description was based. Among the spirit specimens of flowerpeckers in the National Museum of Natural History, Smithsonian Institution, there are spcimens which were partly dissected for examination of the tongue. These specimens, including some of Dicaeum sanguinolentum, D. celebicum, etc., could possibly be part of the material examined by GARDMER. However, I did not locate his specimens of Prionochilus. Whether GARDNER's 86 H. MORIDKA [Jap. J. Ornithol. Vol.40.No.3

description of the Prionochilus tongue is erroneous based on a misidentified specimen or it represents another type of tongue in Prionochilus is a matter of conjecture until the tongue of all six species of the genus has become known, but this is not an important question for the present paper.

MATERIAL EXAMINED

I examined the tongues of four specimens of Prionochilus plateni from Palawan and one specimen of P. olivaceus from Mindanao. These tongues are attached to the trunks which were preserved in alcohol when skins were prepared. Several spirit specimens of P. plateni were examined to ensure the identification of the material. I have also studied the tongue of the following species by opening the bill of spirit specimens and pulling the tongue sideways, but no attempt was made to remove the tongue from the spirit specimen for complete inspection: Dicaeum agile, D. aeruginosum, D. aureolimbatum, D. nigrilore, D. bicolor, D. australe, D. trigonostigma, D. hypoleucum, D. pygmaeum, D. erythroth- orax, D. aeneum, D. sanguinolentum, D. celebicum, D. ignipectus, D. cruentatum, D. trichileum, Melanocharis versteri, Paramythia montium, and Pardalotus punctatus. Two, sometimes three or four, spirit specimens were examined for each species except for Dicaeum agile, D. nigrilore and Paramythia montium, which were represented by a single specimen. The majority of the spirit specimens are in the collection of the National Museum of Natural History, Smithsonian Institution, Washington, DC, and others in the National Science Museum, Tokyo.

DESCRIPTIONS OF THE TONGUE

The two species of Prionochilus (P. plateni and P. olivaceus) agree closely in the structure and shape of the tongue. It is more or less fleshy, relatively thick, shallowly concave on the dorsal surface, and distinctly notched (but not deeply cleft or split) at the tip, with a fine hair-like fimbriation along the anterior third of the lateral margins (the fimbriation is somewhat variable individually and totally lacking in one specimen of P. plateni). As usual of the

Figs. 1-2 Tongues of 1. Prionochilus plateni, 2. Melanocharis versteri. Drawn not to the same scale. September Tongue of Two Species of Prionochilus from the Philippines 1992 ] 87

Figs. 3-12. Tongues of thin-billed species of Dicaeum . 3. D. sanguinolentum, 4. D. cruentatum, 5. D. erythrothorax, 6. D. pygmaeum, 7. D. aeneum, 8. D. hypoleucum , 9. D. trigonostigma, 10. D. nigrilore, 11. D. celebicum, 12. D. australe. Drawn not to the same scale . 88 H.MORIOKA Ornithol.Vol.[Jap.J. 40 No. 3 tongue, there is a row of fringes at the posterior end. In these features the tongue of Prionochilus (Fig. 1) is very similar to that of Melanocharis (Fig. 2), the two differing chiefly in shape (more slender in the former), but is quite unlike those of many species of Dicaeum (described below). In Paramythia the tongue is small for the size of bird, rather flat and thin, and more or less cornified anteriorly, but the tip is not deeply split (see the figure in MAYR & AMADON1947). In both Melanocharis and Paramythia the lateral margins are smooth. The typical tongue of Dicaeum is fleshy only at its posterior half. The anterior half is horny, thin and translucent, and is deeply cleft to form two ribbon-like projections, which are often flat but sometimes semitubular because the edges are curled. The lateral margins are usually smooth (occasionally there is some fimbriation as in Fig. 6), whereas the dorsal surface is more or less deeply concave by curling of the lateral margins or, in a few species, distinctly grooved along the median tine. However, in the specimens examined, the curling of the lataral margins is incomplete and in no case was a tubular structure formed by the curled edges meeting dorsally. I have found this type of tongue in D. pygmaeum, D. erythrothorax, D. aeneum, D. sanguinolentum, D. ignipectus, D. cruentatum, and D. trochileum (Figs. 3-7). The tongues D. hypoleucum (Fig. 8) and D. trigonostigma (Fig. 9) represent a slight modification of the typical one. In these species the tip is cleft into four parts of varied length and shape (some individuals have only two tips) and the semitubular nature of the structure is more pronounced than in the other species. Among Dicaeum the most specialized tongue may be found in D. nigrilore (RAND 1961). In this species (Fig. 10) the tip of the tongue is tripartite, instead of bifid in the typical tongue, and the central element is further subdivided into two parts, resulting in four flat ribbons at the tip (the central pair have an additional short cleft to produce six separate tips). The outer margins of the anterior half are frayed to the tip. My specimen differs from RAND's (1961) description and figure only in minor details, the differences being chiefly in the extent of the marginal fraying and due probably to individual, age, or seasonal variation. To the other extreme, thick-billed species (D. bicolor, D. aureolimbatum, D. aeruginosum) have a tongue which is flat, rather thick and fleshy, and notched at the tip (Figs. 13-15). This condition is much the same as in Melanocharis and Prionochilus, but has hitherto been unknown in Dicaeum.

Figs. 13-15. Tongues of thick-billed species of Dicaeum. 13. D. bicolor, 14. D. aureolimbatum, 15. D. reruginosum. Drawn not to the same scale. D. agile is similar to D. celebicum (Fig. 11). September 1992]Tongue ofTwo Species ofPrionochilus fromthe Philippines 89

Fig. 16 Tongue of Pardalotus punctatus.

The tongue of Pardalotus punctatus is flat, broad, and thin, rather than fleshy, with an irregularily notched tip (Fig. 16).

DISCUSSION

Most flowerpeckers are tiny, restless birds with a small bill. They congregate at flowers, but the exact manner of their feeding and the kind of food taken are usually difficult to observe, due to the small size and the tendency to feed in more or less higher branches. The analysis of stomach contents reveals only part of the actual food habit of these birds since nectar and pollen leave no visible traces in the alimentary tract. Thus, except for a few species, the feeding habits of flowerpeckers are little known and this holds true of Prionochilus. The structure of the tongue may then be used as an indication of their food in the absence of useful information. The tongue of Prionochilus is apparently not adapted for collecting nectar and pollen. Its tonugue can be protruded but little beyond the tip of the stout bill so that it would function mainly for swalling the taken food and regurgitating undesired substances from the mouth. It is assumed that the staple food of Prionochilus is small berries and seeds as in Melanocharis, with insects and spiders also taken frequently or occasionally. Several species of Dicaeum that have a thick bill and a flat, non-split tongue may also be primarily berry and seed feeders. The split, fringed, and semitubular tongues of Dicaeum are probably adapted for an omnivorous feeding habit. These tongues are longer and narrower, and can be protruded to some extent for feeding. They are no doubt capable of collecting nectar and pollen and these items are part of their food, in addition to berries, seeds, insects, spiders, etc. However, none of the flowerpeckers examined possesses a completely tubular tongue like those of sunbirds and spiderhunters, indicating that flowerpeckers are less specialized nectar and pollen eaters. Judging from the structure of the tongue, nectar and pollen would be second in importance to berries, seeds and insects as their food, for otherwise selection favors a tubular tongue. The tubular tongue ofsunbirds and spiderhunters (Nectariniidae), on the other hand, is so slender and thin that it cannot be used for anything but collecting nectar, juice and pollen, and these birds must rely on the action of the bill to catch insects and spiders; berries and seeds could be taken only accidentally (see GARDNER 1925, SCHARNKE 1932, and MAYR & AMADON 1947 for description of the tongue in Nectariniidae; the tongue of Arachnothera longirostra I examined, a sparrow-sized spiderhunter, measured about 60 nun long, 3 mm wide at the base, and only 0.5 mm in diameter of the tube). Among the species of Dicaeum there is a good correlation between bill shape and tongue 90 H. MORIOKA [Jap. J. Ornithol. Vol.40 No. 3

Fig. 17. Stomach of Prionochilusplateni Note that the gizzard is so positioned that easily digestableberries are passed to the intestine without entering the gizzard (see arrow) structure. In general thin-billed species tend to have a typical, split tongue. The fringed and semitubular tongues are found in species with a thin and decurved bill, whereas thick-billed species have a more fleshy, non-split tongue. However, the tongue of thin-billed D. celebicum (Fig. 11) is non-split, flat and fleshy, resembling those of the thick-billed species except for being narrower, in accordance with the narrower mouth cavity. Thin-billed species are thus not always moe specialized for nectar and pollen feeding than thick-billed species (the reverse may not be true; I doubt that there is any thick-billed species that is specialized for nectar feeding). The tongue of D. australe (a thin-billed species, Fig. 12), however, may represent a stage towards specialization. As pointed out by MAYR & AMADON(1947), it is highly probable that the more specialized tongue of Dicaeum could have been evolved from a generalized type found in the primitive and omnivorous flowerpeckers (Melanocharis). Although the tongue of Prionochilus represents the primitive condition, it is not known whether Prionochilus is primitive and related to Melano- charis as sometimes suggested (cf. SALOMONSEN1960). Aside from the developed tenth primary there is little evidence to show that Prionochilus is primitive and more closely related to Malanocharis than to Dicaeum. In this connection it is interesting to find that the stomach of Prionochilus (Fig. 17) is the specialized Dicaeum type, and not a genaralized Melanocharis type, although a relationship between Prionochilus and mistletoe berries is unknown (see DESSELBERGER1931 and MAYR & AMADON1947 for the specialization of the Dicaeum stomach). Presumably Priohochilus and several thick-billed species of Dicaeum have ratained the primitive, unspecialized tongue in the absence of the need for nectar feeding.

ACKNOWLEDGMENTS I am greatly indebted to the authorities of the National Museum of Natural History, Smithsonian Institution, and Dr. Richard L. Zusi for allowingme to examinethe spirit specimensof flowerpeckersused in the present study. September 1992] Tongue of Two Species of Prionochilus from the Philippines 91

SUMMARY

The tongues of Prionochilus plateni and P. olivaceus were examined and redescribed. These tongues are very similar to the primitive condition found in Melanocharis. Implications of these facts are discussed. Prionochilus and several thick-billed species of Dicaeum probably retained the primitive, unspecialized tongue in the absence of a need for nectar feeding, while they are typical Dicaeum in many respects.

摘 要

フ ィ リ ッ ピ ン産Pionochilus属 の2種 の ハ ナ ド リ類 の 舌 を 調 査 し,再 記 載 し た.こ の 舌(Fig.1)は, 比 格 的 肉 質 で 厚 く,上 面 は む し ろ 平 ら で,先 端 が さ さ くれ て い る.ハ ナ ド リ科 の 中 で は 原 始 的 なMela-

属 の 舌(Fig.2)に 構 造 ・形 状 と も き わ め て よ く似 て い る. nocharis

い っ ぽ う,Dicaeum属 の ハ ナ ド リ類 の 多 く の 種 で は,舌 の 前 半 部 は薄 い 角 質 で,2部 に 分 か れ,ひ も

状 な い し半 管 状 の 突 起 を 形 成 し て い る(Figs.3-7).も っ と も 特 殊 化 の 進 ん だDicaeum nigriloreの 舌

(Fig.10)は,先 端 が6部 に 分 か れ,そ の 外 縁 に は ブ ラ シ状 の 毛 が あ る.し か し,Dicaeum属 の 中 で も の 厚 い 種 は,Melanochnris型 の 単 純 な 舌 を 持 つ(Figs.13-15). 嘴

一 般 に,よ り特 殊 化 し た 型 の 舌 は嘴 の 細 長 いDicaeum属 の ハ ナ ド リ類 に 見 出 さ れ る が,例 え ばDi- caeum celebicumの よ う に,嘴 が 細 長 くて もMelonocharis型 の 舌 を 持 っ も の が あ る(Fig.11).

Prionochilus属 の 舌 は,明 らか に 花 蜜 や 花 粉 の 採 食 に 適 応 して い な い.し た が っ て,彼 ら の 食 性 は, Melanacharis属 の 場 合 と 同 様 に,小 果 実 ・漿 果 ・種 子 な ど を 主 と し,昆 虫 類 や ク モ 類 も食 べ て い る と

考 え ら れ る.

舌 の 形 態 か ら み る と,Dicaeum属 の 多 く の 種 は 花 蜜 食 に 適 応 して い る.し か し,ハ ナ ド リ科 に は タ イヨゥチョウ科のように完全に管状の舌を持つものはなく,花 蜜や花粉は彼 らの食物の比較的小部分

を 占 め る に す ぎ な い で あ ろ う.も し花 蜜 や 花 粉 が 主 食 で あ る な ら ば,管 状 の 舌 が 進 化 した に 違 い な い.

ハ ナ ドリ 科 で は,舌 は 果 実 食 のMelanocharis型 か ら,花 蜜 食 も 可 能 な 先 端 の 分 か れ た も の(Dicaeum

属 の 大 部 分 の 種)へ と 進 化 し た が,Prionochilus属 やDicaeum属 の 一 部 の 種 で は,何 らか の 理 由 で 花 蜜

食 の 必 要 が な か っ た た め に,原 始 的 な 形 態 の 舌 が そ の ま ま 残 っ た と推 察 さ れ る.

LITERATURE CITED

DESSELBERGER,H., 1931. Der Verdauungskanal der Dicaeiden nach Gestalt and Funktion. J. Ornithol. 79:353-370. GARDNER,L. L., 1925. The adaptive modifications and the taxonomic value of the tongue in birds. Proc. US. Nat. Mus. 67(19): 1-49. MAYR, E., & AMADON,D. 1947. A review of the Dicaeidae. Am. Mus. Novitates (1360): 1-32. RAND, A. L., 1961. The tongue and nest of certain flowerpeckers (Ayes: Dicaeidae). Fieldiana (Zool.) 39: 581-587. 1967. The flower-adapted tongue of a Timaliinae bird and its implications. Fieldiana (Zool.) 51: 53- 61. SCHARNRE,H., 1932. Ueber den Ban der Zunge der Nectariniidae, Promeropidae and Drepanididae nebst Bemerkungen zur Systematik der blutenbesuchenden Passeres. J. Ornithol. 80: 114-123. SALOMONSEN,F., 1960. Notes on flowerpeckers (Ayes, Dicaeidae). 1. The genera Melanocharis, Rhamphocharis, and Prionochilus.. Am. Mus. Novitates (1990): 1-28.

(Accepted 15 March 1992)