NAT. NAT. HIST. BUL L. SIAM Soc. 46: 27 -4 2， 1998

THE BREEDING BIOLOGY OF THE ASIAN PARADISE FLYCATCHER TERPSIPHONE PARADISI IN KHAO PRA-BANG KHRAM WILDLIFE SANCTUARY ，SOUTHERN THAILAND

Taku Taku Mizuta 1

ABSTRACT

The breeding biology of the Asian Paradise Flycatcher Terpsiphone paradisi was studied at at Bang Tieo area ， in Kh ao Pr a-Bang Kh ram Wildlife Sanctuary ，Kr abi Province ，southern 百lailand ，合 om March-July of 1993 and 1994. Th ree types of male were observed in 白is study area: area: white-plumaged males long with tails (WL) ，rufous-plumaged m a] es long with tails (RL) ，

組 d rufous-plumaged males with short tails (RS). Each male type was found breeding. Th ey started started their breeding attempts in mid-March. Both males and fem a] es incubated eggs and fed 血eyoung. V 訂 ious p紅白 of nests were measured and there were no differences in characteristics among nests of these three male types. RS males started their breeding attempts later 出an both long-tailed long-tailed males. 百le clutch size of females paired with RS males was smaller than 出at of fem a] es p町吋 wi 出 long-tailed males. Wh en the brood size was 由民e，出 e body weight of

nestlings nestlings of long 司 tailed males was found to be heavi 紅白an that of RS males. RS m a] es had significantly significantly smaller wing and tail length ， and lighter body weight ， than long-tailed males. 百lese findings indicate 出at RS males were younger 出 an long-tailed males. 百le elongated central central tail feathers of RL males were significantly shorter than those of WL males. It was thought 出at WL males are older than RL males ， but the reason for colour dimorphism in 曲目 species species is still unknown.

INTRODUCTION

The genus Terps 伊hone (Mon 訂 chinae) is widespread in Asia and Af rica. Th ere are 12 12 species in this genus ，and in most species males have longtail feathers and plumage dimo 中hism (PETERS ， 1986). The Asian Paradise Flycatcher Terpsiphone paradisi is distributed from Russian Turkestan east through India ， Sri Lanka ， north-central and eastem China ，Korea ，and Indochina ，south and east through the Malay archipelago to Sumba and Alor islands (MA YR & COTTRELL ， 1986). It inhabits evergreen and mixed deciduous forests ，secondary growth ， from the plains to 1，500 m (LEKAGUL & ROUND ， 1991). These f1 ycatchers feed mainly by hawking insects in the air ，and they are common residents in southem Th ailand.τ 'h ere are are 16 subspecies in these geographical areas (OWEN ， 1963; PE 百 RS ， 1986). OWEN (1963) determined the geographical trends by examining relative frequency of occu 町 'ence of rufous and white males by examining the specimen occuηence in several museums. The ratio of occu 町 ence of rufous and white long-tailed male specimens appe 紅 s to vary geographically.

1 Department of Zoology ，Faculty of Science ， Kyoto University ，Ki tashirakawa ・oiwakecho ， Sakyo-Kyoto 60 6- 01 ，Japan. Paper Paper presented at the Second Intemational Asian Hombill Workshop ，Bangkok ，13 ー 15 April 1996.

27 27 28 T AKU MJZUTA

(c) + • -Tracks

SO Om Scale

Figure I. Maps of the study area. (a) Thail and. (b) Kh ao Pra Bang Khram Wi ldlife Sanctuary. (c) Bang Tieo study area. BREEDING BIOLOGY OF THE ASIAN PARADISE FLYCATCHER 29

Adult Adult males have either predomin 叩 tly bright rufous upper p紅白 or predominantly white. In In both colour types of adult males ， the two central tail feathers are greatly elongated 叩 d form streamers. Long-tailed rufous males are generally devoid of shaft streaks on the wing and tail feathers ，which in white males are blac k. Y oung males 紅 e rufous and have short tails. tails. Females also have rufous upper p紅 ts and short tails (OWEN ， 1963). Some authors (e.g. (e.g. SALOMONSEN ， 1933) consider rufous and white males as distinct colour forms or ccolour ccolour phasese ， while others (e.g. RILEY ， 1938) consider the rufous males to be young adults adults and the white males full adults. Wi 血 respect to 白is problem ，OWEN (1963) suggested some explanations. First ，two types of plumage colour of long-tailed males could be a genetic genetic polymo 中hism; second ， rufous and white males could be different ages ， the long- tailed tailed rufous males moulting into white plumage when they grow older; third ， each of the above above explanations could be true for some species; and fourth ， the two differently coloured males males could in fact be two sibling species. No one has ever studied the plumage dimorphism of 白is P釘 adise flycatcher in the wild ， so reasons for this dimorphism 訂 e not yet exactly known. In 出is s加 dy ，1 conducted field rese 紅 'ch in a forest in southem τ'h ailand ，and repo 口 some basic information about the breeding biology of the Asian Paradise Fl ycatcher ， particularly particularly with respect to the colour dimorphism in males.

STUDY AREA

The field work was conducted at Bang Tieo ， in Kh ao Pra- Bang Kh ram Wildlife S叩 ctu ぽ y，Kr abi Pr ovince ，southem Thailand (8 0 10"N and 98 0 80"E ，maximum elevation of of 650 m) (ROUND & TREESUCON ， 1992) (Fig. 1) from March-July in 1993 阻 d 1994. 百le vegetation vegetation is mainly constituted by small remn 叩 t patches of lowland rain fores t. More than than 290 species of birds have been recorded in the 紅白. Most of the plant species 紅 e pioneer pioneer or secondary grow 出 species due to the practice of wood cutting 30 or 40 years ago (ROUND & TREESUCON ， 1992). There is a network of nature trails in 血is study area

(Fig. (Fig. 1)， so it is rather easy to walk 釘 ound the fores t. There are many small stre 創 ns in the the forest ， but some of them are dry when rain stops. In 白is 紅 ea ， the rainy season starts in in April and ends in October or November. However ，it does not rain every day ， but once it it does ，it rains very hard. 百le 合y season is the rest of the year. The breeding season of of most birds here is synchronised with the rainy season.

METHODS

Identification Identification of males and females.- Males were classified by their plumage colour colour and taillength. White and rufous-plumaged males with long tails were referred to as as “WL male" and “RL male" resp 田 tively ， whereas rufous-plumaged males with short tails tails were refe 汀 ed as “RS male" (Fig. 2). In some cases ，WL males and RL males were lumped toge 出er and 紅 e called “Long-tailed males". All types of males had bright-blue

eye-rings. eye-rings. Females resembled RS males ， but their eye 旬 rings were not as bright as of males and were used to distinguish between RS males and females. 30 TAKU MIZUTA

Nest Nest searching.-I hiked in 出e forest 仕om 060 0- 1200 hours and from 1600 ー1800 hours hours almost eveηday ，from early March to mid-July. 1 walked along the nature trails in in the study area searching for nests of the birds or if 1 heard a call of the flycatcher ，1 follow follow the bird. If it was a male ， 1 identified the type (mentioned above). When 1 found a nest ，1 inspected its contents. Once nests were found ，1 observed each one every day. Measurements of a nest.- Various parts of nests and nest 紅白s were measured after the the chicks fledged or disappeared before fledging. Measurements included the nest height above above the ground ，diameters of nest (inner and outer) ， depth of nest ， nest tree height ， trunk diameters diameters of nest 住ee (at the base ，breast height ， and at the nest) ，number of forks at the nest ，number of trees around the nest (within 10 x 10 m). These trees were classified by three three categories of diameters at breast height (DBH). Th e species of nest 住民were identified. Breeding stages.-I checked the morph of the male at each nes t. All nests were inspected inspected every day to certify the dates of laying ， hatching and fledging. The breeding stages stages of the nest were divided into three periods: nest building ，incubation and brooding. The nest building period is defined as when the nest was obviously under construction ， and incubation incubation and brooding periods 紅 e defined as the times when there were eggs or chicks in in the nest ，respectively. The incubation period is determined from the day on which the first first egg was laid until one day before hatching. The brooding period is determined from the the day on which the first egg hatched until fledging. Wh en 1 found a nest with no egg

an d/ or chick ， and eggs did not appe 訂 aft 巴r more than two weeks ，1 considered the nest as being being predated and abandoned. 1 also recorded the number of eggs an d/ or chicks in the nes t. Chicks were weighed when they were 9 days old ， one day before fledging (see below). below). Nest observation.- To minimise disturbances to the birds' activities during 白e brooding brooding period ，all nests were observed from an observation hide 10-20 m from the nest using using a spotting scope (30x) or a pair of binoculars (7x). 1 divided the brooding period into into two parts: early brooding period (from hatching to 5 days old) and late brooding period period (from 6 days old to fledgling). To observe late brooding periods ，1 spent 90 minutes at at each nest during the morr 由19S of 1994. Total observation time for 14 nests was 1，260 minutes. minutes. Frequency of feeding behaviour was recorded during the 90 minute periods. Data Data of all three male types were pooled due to insignificant differences of feeding rates. 1 then classified the data on feeding rates by brood sizes. Capture ， linear measurement and marking of birds.- Adults: Adult birds were captured captured in front of the nest using mist nets. This was performed during the late brooding period ， otherwise it was too risky for nest abandonmen t. For individual identification ，a captured captured bird was marked with an aluminium ring on the right leg and four colour rings (two (two rings on each leg). Then the following measurements were taken: length of wing ， tail ，tarsus and bill and body weigh t. Tail length was taken as the length of the longest of of the two central feathers. Chicks: Chicks were marked at the nest with an aluminium ring ring and four colour rings in the same way as adults when they were 9 days old.τbe criterion criterion used in selection of chicks at this age was tarsus length ， as the leg must be long enough to c紅 ry all the rings. Body weights of chicks were also taken at this age. To prevent prevent parent birds from being disturbed ， marking and weighing were done during their absence. absence. BREEOING BIOLOGY OF THE AS lA N PARA Dl SE FL YCATCHER 31

RESULTS

Breeding Biology

Types Types 01 breeding male Thre 巴 types of mal 巴 Asian Paradise F lycatch 巴r were obs 巴rv 巴d in th 巴 studyar 巴a，na ll1巴 Iy ， WL ，RL and RS m al es (F ig. 2). Each male type was found in bre 巴ding cond iti on. No whit 巴 mal 巴 with a sh ort tail was observed.

Characteristics Characteristics 01 nests A total of 31 and 57 n巴sts ，in cludin g abandon 巴d or pr 巴dated n巴sts ，wer 巴 found in 1993 and 1994 ，resp 巴ctively. Measure ll1 ents of th 巴 n巴sts and n巴st t1 ・巴 es 倒・巴 shown in Table 1. Nests Nests were built on a fork of the tree . T he h巴ight of the nests was less than 2 1l1. Nests were open with a deep bowl-shap 巴. The nest materials were finely int 巴r1 aced twigs and dri 巴d grass blades ，r巴 inforced with 1l1 0SS 巴s and spider webs. Fin 巴 fibre-like mater ・ials lin 巴d th 巴 insid 巴 of th 巴 nest ，on which the eggs w 巴re laid. Ther 巴 were no differ 巴nces 111 n巴st characteristics characteristics among nests of the thre 巴 types of ll1 ales (Kruska l-Wa lli s tes t: all n .s.) (Tab le 1). The n巴st trees were small with a m 巴an h巴ight of less than 3 m (Tabl 巴 1) . In 1993 ，25 out of 31 species of nest tr 巴巴 s wer 巴 identified (Tab le 2). Some tr 巴es were used more commonly than othe rs: six n巴sts were built on Schun αwa /liclui ，tlu ・巴 e on Phy l/ anthus oxyphyllus oxyphyllus two each on Vatic α Lowii and Ma Ll otus subpe Lt atus; J2 other sp 巴cles w 巴re used only only once .

Long tailed male

WLmale RL male RS male Female

Figure Figure 2. Th ree types of male and fema le Asian Parad ise Flycalchc l Table Table 1. Measurements of nests and nest trees obtained in two years' sωdy. bh = breast height ，DBH = diameter at breast heigh t. tトa品必

Nest Nest charac 伽istics Nest tree characteristics Numberof N醐 .ber of trees in a plot Mal etype height diameter (佃) depth 民i邑t 回目 kdiame 伽(四) forks at (lOxl 伽)制剛山崎 L泊size やm) Ulsg diameter Short diameter (cm) やm) at base 紙bh at n鎚t 血.e nest clas 蜘 byDBH (cm) ter Ou ter Inn er 伽旬r 加 er <5 5- 30 >30

All All males Ran ge 11 8- 297 7. 8- 10 .3 6.1 -7 .7 6. 4-8 .5 4.ι6.5 2. 6-3. 8 157-389 1. 0ー2.9 0. 6-1. 7 0ふ1. 6 2-3 3- 112 1-28 0-4 ::44 N::44 M 醐 170.6 8.6 6.8 7.3 5.6 3.1 240.2 1. 9 1.1 0.9 2.1 59.2 11. 3 1.1 S.D. S.D. 48 .7 0.5 0.4 0.6 0.3 0.3 55 .3 0.4 0.4 0.3 0.3 22.5 5.8 1.3

Uls g. 凶led Ran ge 118 ー296 7. 8- 10 .3 6.1 -7 .7 6，叫 .5 4. 8-6 .5 2.6 ー18 157-389 1ル2.9 0.7-1. 7 0. 5-1. 6 2-3 1 8- 112 4- 28 0-4 males males M 阻 167 8.6 6.8 7.2 5.6 3. 1 23 9.4 1. 9 1.1 0.9 2.1 61. 7 1 2. 0 N=30 N=30 S.D. 50.4 0.5 0.5 0.5 0.4 0.4 59 .7 0.4 0.4 0.3 0.3 20 .7 6.0 1.3 →長c WLmale WLmale Ran ge 118 ー296 8.1 -9 .5 6.1 -7 .7 6.' 叫.5 5. 3-6 .5 2. 6-3. 8 157-389 1. 0ー2.3 0.7-1. 7 0 .7-1. 6 2-3 1889 4- 18 0- 3 N=12 N=12 M 悶 168.9 8.6 6.7 7.3 5.5 3. 3 244.6 1. 9 1.1 2.1 57 .1 10.6 0.9 .D. S.D. 46 .3 0.5 0.6 0.7 0.3 0.4 55 .4 0.3 0.3 0.3 0.3 19.9 4.5 1.1

阻， male Ran ge 11 8- 297 7. 8- 10 .3 6. 3- 7.6 6.. 牛 7.9 4. 8-6 .2 2.ι 3. 5 157-350 1.1 -2.9 0.7-1. 5 。ふ1. 4 2-3 2 6- 112 5- 28 0-4 N=18 N=18 M 聞 165 .7 8.6 6.9 7.2 5.6 3. 0 235.9 1. 9 1.1 0.8 2.1 ω.8 12.9 1.1 S.D. S.D. 54 .3 0.6 0.4 0.5 0.4 0.3 63. 8 0.4 0.4 0.3 0.3 21. 2 6 .7 1. 5

郎 maI e Ran ge 125- 274 8.1-8 .9 6.1 -7.0 6. 9-8 .4 5. 4-6.3 2. 6-3. 5 178-342 1.1 -2.5 0. 6-1. 7 O.ι 1. 3 2-3 3- 96 1-22 0-4 N=14 N=14 M 聞 178.4 8.6 6.6 7.7 5.7 3 24 1. 9 1. 9 1.1 1. 0 2 .1 53. 7 9.9 1.3 S.D. S.D. 45.6 0.3 0.3 0.4 0.3 0.3 46.5 0.5 0.3 0.2 0.3 26.0 5.6 1. 3 BREEDING BIOLOGY OF THE ASIAN PARADISE FLYCATCHER 33

Table Table 2. Families and species of 血e nest 紅白s of Asian Paradise Fl ycatchers studied in 1993 1993 at Bang Tieo. ? = predated or abandon 吋 nes t. * RL m a1 e replaced WL m a1 e.

Nest Nest No. Type of Nest tree male Family Scientific name

NNNNNNNNNMMMMMMMMMMMMMMMMMMMMMmm官 山?机机ヲ机紅机机?紅悶悶紅郎郎9??郎乱郎ワ机守???99? i Pandaceae Galearia 舟lve 司A Euphorbiaceae Euphorbiaceae Phyllanthus 0巧， phyllus 今 3a Guttiferae Crat o. 勾，z on maingayi 斗 Rubiaceae Ixora cibdela ζJrO Th eaceae Schima wallichii Dipterocarpaceae Dipterocarpaceae Vatica lowii 『 。F Euphorbiaceae Phyllanthus oxyphyllus onynu'A Theaceae Schima wallichii Papilionaceae Papilionaceae Derris sp. Violaceae Violaceae Rinorea sp. Apocynaceae Wrightia dubia 司 L Theaceae Schima wallichii 勾3a Rubiaceae Rubiaceae P勾，chotria sp. 斗 Euphorbiaceae Glochidion sp. εJζU Euphorbiaceae Euphorbiaceae Baccaurea parviflora Euphorbiaceae Euphorbiaceae Phyllanthus o. 巧IfJ hyllus 司 Ioon Unknown Di pterocarpaceae Vatica lowii ヲ Unknown nu't Myrtaceae Myrtaceae Syzygium operculata Eupho 出iaceae Mallotus subpeltatus 司， B D i1l eniaceae Dillenia obovata 弓 3a Verbenaceae Vitex pinnata 斗 Euphorbiaceae Aporosa aurea H3ζU Th eaceae Schima wallichii Theaceae Theaceae Schima wallichii 『 Foonynu'i Myrtaceae Syzygium oleina Euphorbiaceae Euphorbiaceae Mallotus subpeltatus Theaceae Theaceae Schima wallichii Rutaceae Rutaceae Micromelum minutum Th eaceae Eurya acuminata

* RL m a) e replaced WL male 34 TAKU MrZUTA

Egg laying and incubation The breeding season of the Asian Paradise Fl ycatcher st 副 s in early March and ends in in mid-July. Dates of first egg laying and fledging of chicks 紅 e shown in Figs. 3a 加 d 3b ，respectively. The distribution of egg laying date and fledging date did not differ between between the two years when nests of the three types of males were pooled (Mann- Whitney U-test: U-test: Z = -0 .52 ，n = 16 ，30; nふ). There was a significant difference among dates of first egg egg laying of the three types of male (Kru skal-Wallis test: H = 14.67 ，n = 12 ，21 ， 13; p = 0.0007). WL male pairs laid first eggs earlier than did RS male pairs (Mann-Whitney U-tes t: Z = -3.16 ，n = 12 ， 13; p = 0.0016) ， and RL male pairs also laid frrst eggs earlier than than RS male pairs (Mann-Whitney U-tes t: Z = -3.35 ，n = 21 ， 13; p = 0.0008). When the the first egg-laying dates of long-tailed males were pooled ， they proved to be significantly earlier earlier than those of RS males (Mann-Whitney U-tes t: Z = -3.72 ，n = 33 ， 13; p = 0.0002). There There was no significant difference in date of first egg laying between pairs of the two long-tailed long-tailed male types (Mann-Whitney U-tes t: Z = -1.1 6，n = 12 ， 21; nふ). Th ere was also also no difference in fledging dat 巴 between pairs of the three male types. This might be affected affected by a too small sample size. The incubation period lasted 14 ー16 days (mean 14.90 :!: 0.74 s.d. ，n = 10) and nestling period period lasted 9-12 days (mean 1O.4 2:!: 1. 00 s.d. ，n = 12). When 1 found a nest for which 1 was unable to confirm egg laying and hatching dates ，1 determined the laying date by counting counting 15 days backwards from hatching and the hatching date by counting 10 days backwards backwards from fledging.

Clutch Clutch size and young The numbers of eggs and young are shown in Table 3. The clutch size was usually three three and sometimes two (mean 2.79 :!: 0.4 1 s.d. ，n = 39). The eggs/chicks decreased in number as the br 田 ding period proceeded due to cracking an d/ or crushing. The number of of eggs and nestlings of long-tailed male pairs were significantly larger than those of RS

male male pairs (Mann 聞 Whitney U-tes t: number of eggs: Z =ー 2.15 ，n = 27 ， 12; p = 0.0313; number of nestlings: Z = -2.09 ，n = 28 ， 12; p = 0.0364). There was no significant difference difference in the number of eggs and the young between "¥¥屯 and RL male pairs (Mann- Whitney Whitney U-tes t: number of eggs: Z = -1.1 1，n = 10 ， 17; p = 0.0313; number of nestlings: Z = -0 .57 ，n = 10 ，18; p = 0.0364) (Table 3; Fig. 4). Body weight of the young differed with with brood size. When the brood size was two ，body weights of the young were slightly heavier heavier than when the brood size was three ， but the difference was not significant (Mann- Whitney Whitney U-test: Z =ー 1.4 4，n = 20 ， 27; n.s.). All types of male pairs could raise young with with similar body weight when the brood size was two (Kru skal-Wa 11i s tes t: H = 1. 49 ， n = 4，10 ， 6; nふ) (Fig. 5). In contrast ，when the brood size was three ，body weights of the the young differed with male type. Long-tailed male pairs raised significantly heavier young than RS male pairs (Mann-Whitney U-tes t: Z = -2.16 ，n = 21 ， 6; p = 0.0306) (Fig. 5). 5). Th ese results indicate 出at RS male pairs could not raise the young well if the brood size size exceeded two. However ， sample sizes were too small to make a definite conclusion

Behaviours Behaviours during late brooding period The total feeding rate by both p紅 ents increased with brood size (Fig. 6). Males contributed contributed significantly to 出 e increases of feeding rate in accordance with the brood size BREED lN G BIOLOGY OF THE ASIAN PARADISE FL YCATCHER 35

10 10 WL male 9 口 圏 RL male 8 RS male 幻自ロ』7 口 6 O ωS53Z』 5 4

3 2 。1 5 10 15 20 25 30 3540 45 50 55 60 65 70 75 80 85 Date Date (1= March 22)

Figur 巴 3a. Two yea r poo led data of firsl egg la yin g date s of lhr ee mal e lypes (1993 and 1994)

7 口WL male

己的6 圏 RL male ω ロ』5 RS male O 』 4 円以ロ同ロω 3 Z 2 。1 5 10 15 20 25 30 35 40 45 50 55 60 Date (1= April 26)

Fig ur e 3b. T wo year poo l ed data of ll ed呂ing dates of lhree ma le ty pes (1993 and ( 994). 36 36 TAKU MIZUTA

Table Table 3. Number of eggs ，nestlings and fledglings of three male types. Data pooled from two years' sωdy (1 993-1994).

Male type Number of eggs eggs nestlings fledglings

All All males Range 2-3 0- 3 。-3 Mean 2.79 2.50 1. 67 S.D. S.D. 0.4 1 0.85 1. 28 N 39 40 42 Long-tailed Long-tailed Range 2-3 0- 3 ι3 males males Mean 2.89 2.68 1. 83 S.D. S.D. 0.32 0.67 1. 23 N 27 28 30 WL male Range 2-3 0- 3 ι3 Mean 2.80 2.50 2.00 S.D. S.D. 0.4 2 0.97 1.1 8 N 10 10 11 RL male Range 2-3 2-3 0- 3 Mean 2.94 2.78 1.7 4 S.D. S.D. 0.24 0.4 3 1. 28 N 17 18 19 RS male Range 2-3 。-3 0- 3 Mean 2.58 2.08 1. 25 S.D. S.D. 0.51 1. 08 1. 36 N 12 12 12

(Kru skal ・Wallis tes t: H = 7.33 ，n = 3，5 ，6; p = 0.0256) ， whereas feeding rate by females did did not increase sig 凶ficantly with brood size (Kru skal-Wallis tes t: H = 1. 42 ，n = 3，5 ，6; nふ) (Fig. 6). So the total feeding rate per nes t1 ing did not differ among different brood sizes. sizes. Increases in total feeding rate depended on the male feeding rate.

Size Size of Males and Females

Body size Males: Males: Wing ，tail ，tarsus and bil1 1engths ， and body weight of three types of male are shown in Table 4. The tail of WL male was significantly longer th 叩 that of RL male (Mann-Whitney (Mann-Whitney U-test: Z = -2.87; P = 0.0041) (Table 4). Long-tailed males had significantly significantly longer wings and tails ， and heavier bodies ，th 叩 RS males (Mann-Whitney U- test: test: wing length: Z = -3.65 ，P = 0.0003; taillength: Z = -3.63 ，P = 0.0003; body weigh t: Z = -2.04 ，P = 0.0418 ，n = 19 ， 6) (Table 4). BREEOING BIOLOGY OF TH E ASIAN PARADISE FLYCATCHER 37

4 米 「一一一

3 ω∞∞ N = ll ω 』。』 2 ω 門出田口 1 Z 。 aguωω 3

ロ』。』 2 ωAEコ 1 Z 。 as- ∞℃3 ロ刷。』ω 2 来 ω Mann - 門戸同 HEZ1 Whitney U test 。 p

Figure Figure 4. Number of eggs and yo ung in nests of thr ee ma le typ es 38 38 T AKU M IZUTA

17 口 WLm 枇

16 関 RLm 山 (∞) 固 RS male

~ 也 15 ω 主 〉、 ち 14 凶

13 来 Mann -Whitney U test P く 0 .05

12 3 youngs 2 youngs

Figllr e 5. Body we ig ht of ne sLl ing s of thre e l1l al e typ es

N=3 N=5 N=6 WL male =O WL male =l WL male=2 RL mal e=2 RL mal e=2 RL male =2 RS mal e= l RS mal e=2 RS male=2 20 20 |圏 Male 口 Female

J Jロt4〉コ J 、 15 15ECロ L

ーナ Tota l/ chick 斗Z4 10 10~ 心可 0.0 0.0 。

・ロ 伊司 1) ロr1) 5 5 ℃弘q4J 4J 、〈 r・3 。 。 1 2 3 Brood size

Figure Figure 6. Two y巴ar poo l ed data 0 1' fr eq ll en cy of feeding by l1l ale and fe l1l ale，wit h different brood s izes (1993 to to 1994) BREEDING BIOLOGY OF THE ASIAN PARADISE FLYCATCHER 39

Table Table 4. Body measurements of the three male types. Data is pooled 仕om two years' study study (1 993-1994).

Male type Wing(mm) Tail(mm) Tarsus (mm) B il1 (mm) Body weight(g)

Al l males Range 86.0 ー95. 。90 .5 -310.0 18.3-20.8 19.2-23.3 19.1-22.9 (N = 25) Mean 91. 76 206.92 19.38 21. 88 20.88 S.D. S.D. 2.6 70.26 0.55 1 1.1 3

Long-tailed Long-tailed Range 9 1.0- 95.0 197-310 18 .3 -20.8 19.2-23 .3 19.ι22.9 males males Mean 92.95 242.68 19 .4 3 2 1. 89 21.1 6 (N = 19) S.D. 1. 41 30.78 0.55 1. 07 1.1

WL male Range 91. 5-95.0 240 ー310 18.3-19.8 21. 3-22 .4 19.ι22.5 (N = 5) Mean 92.8 280 19.2 21. 62 21. 4 S.D. S.D. 1. 44 25.85 0.55 0.79 1.1 7

RL male Range 91 ル95.0 197-266 18.8-20.8 19.2-23 .3 19.7-22.9 (N = 14) Mean 93 229.36 19 .5 1 2 1. 99 2 1. 07 S.D. S.D. 1. 45 19 .4 9 0.54 1.1 7 1.1

RS male Range 86.0 ー90.5 90.5-104.0 18.3-20.0 20.9-22.7 19.1-20.9 (N = 6) Mean 88 93.67 19.22 21. 83 20.03 S.D. S.D. 1. 73 5.79 0.59 0.79 0.78

Females: Females: Wing ，tail ，t紅 sus and bil1 1engths ， and body weight of females which p誼 ed with with the three types of males were measured. There were no differences 創 nong these females. females. Although wing length and body weight of females paired with RS males were slightly slightly shorter and lighter 血an those of females paired with long-tailed males ， the differences were not sig 凶白cant.

Changing 01 plumage colouration and pairing Two RL males ringed in 1993 were recovered in 1994. Al though one retained the RL plumage to breed in 1994 ， the other was found to have changed to the white plumage between between seasons. One female was seen to change mates between the two breeding seasons ，pairing with a WL male in 1993 組 da RL male in 1994.

DISCUSSION

τ'h e breeding biology of the Asian Paradise Fl ycatcher had not been well studied. This This study revealed some differences between long-tailed males and RS males in body 40 TAKU MIZUTA

characteristics characteristics and in breeding. RS males had smaller body size (wing leng 血叩d body weight) weight) than long-tailed males. RS males started their breeding attempts later ，and the clutch clutch sizes of their mates were smaller than those of the long-tailed males. This is similar to to the younger Great Tits Parus major ， which lay eggs a little later ， lay smaller clutches or or are less successful in raising their young (KLUIJVER ， 1951; PER 即 NS ，1965). 百1U S，1 am convinced 白at the RS male Asian Paradise Fl ycatchers are a younger age group 出an the long-tailed long-tailed males. 百lI s is due to the that fact 1 have never seen 加 y white-plumaged ，short- tailed tailed males. Moreover ， the plumage of all nestlings was rufous ，indicating that RS males are are probably yearlings. 1 could not find any differences in frequency of feeding among 出e 由民e types of males ， but it is possible that the difference in feeding rate between long-tailed males and RS males was a consequence of the larger clutch size. RS males may have had less ability in in raising young than long-tailed males whose clutch sizes were larger. It seems that 出e increase increase in number of nestlings from two to three may not cost WL males and RL males much ， but might considerably cost the RS males. Generally in birds ，body weight of the young tends to decrease with increasing brood size (KLOMP ， 1970). In the present study ， the the brood size of the Asian Paradise Fl ycatcher varied from one to three ， and mean body weight weight of the young when the brood was three was only slightly lighter 出an when it was two. two. 1 could not specifically determine whether the difference in feeding rate was really influenced influenced by the number of nestlings or by the male type ， because the data were pooled due due to small s創 nple size. It is possible that the low feeding rate for brood size of one was affected affected by RS males ， and a high feeding rate in brood size of three was affected by WL 組 d RL males. Another explanation is that when 出e number of nestlings is small ， males skimp ， but they are eager to feed when the number of nestlings is large ，reg 紅 dless of male type. type. My explanation for the two different colour morphs in males of 白is species has not yet yet been proven. 1 could find a difference only in taillength between WL males and RL males ， and 1 found one RL male that tumed into a WL male. From these findings ，1 was convinced 血at WL males are older 白血 RL males. According to my observations on change change of plumage colouration ， the sibling species hypothesis proposed by OWEN (1963) can can be challenged. 1 would hypothesize as follows: ye 紅'li ng males possess rufous plumage with with short tail feathers and resemble females. Th en ，by the second year of age the long tail tail feathers develop ， although rufous plumage is retained. Later the rufous plumage may t町 n to white ，although it is unknown at which age the plumage begins to change. Moreover ， it it is unknown whether the change of colour from rufous to white occurs in all males or only only in some individuals genetically predisposed to change (Fig. 7). To investigate the adaptive adaptive significance of the plumage dimorphism ，it is necess 紅 Y to collect data with sufficient sufficient sample sizes of breeding behaviour ，reproductive success ，feeding rate and foraging efficiency efficiency of individual males. BREEDING BIOLOGY OF THE ASIAN PARADISE FLYCATCHER 41

'0 -0 i

l v R?\ ------t------t ------t------V ------Chick

0 0 X

Figure 7. Four possible hypotheses of development of plumage in male Asian Paradise Flycatchers.

ACKNOWLEDGMENTS

I am most grateful to Dr. Satoshi Yamagishi for his kind advice, cntiCISm and encouragement throughout this study and comments on the manuscript. I am indebted to Dr. Pilai Poonswad and members of the Hornbill Project Thailand for their kind hospitality; without their help this study would have been difficult. I am especially grateful to Dr. P. Poonswad, not only for co-operation in the field research, but also for her comments, suggestions and corrections of my English in this paper. I am also grateful to Mr. Philip D. Round, Mr. Uthai Treesucon and all the member of Khao Nor Chuchi Lowland Forest Project in Thailand, for their co-operation during the field research. They made my study and life in the field much easier and pleasant. They also provided me with comfortable accommodation in the study area. I would like to thank Mr. Vicharn Iadthong for help on identification of species of nest trees of Asian Paradise Flycatchers. I also thank Mr. Tasanet Petkhong, the chief of Khao Pra-Bang Khram Wildlife Sanctuary, who allowed me to conduct the research in this area. I am grateful to Dr. Takeo Kawamichi and Dr. Masanori Kohda for their valuable comments and criticism. I am also grateful to Dr. Masanobu Hotta, Dr. Chang-Hoe Kim, Mr. Isao Nishiumi and other members of the Laboratory of Animal Sociology of Osaka City University for their comments through many discussions. 42 TAKU MIZUTA

RE 回 RENCES

KLoMP ， H. 1970. Th e Determination of clutch-siz 巴 in Birds. Ardea 58: 1-124. KL 叩 VER ， H. N. 195 1.百 le population ecology of the Gre at Ti t. Ardea 39: 1-135. LE 臥 G 札， B. AND P.D. ROUND 199 1. A Guide to the Birds 01 Thailand. Saha Karn Bhaet Co. ，Ltd. ，百 lailand. MAYR ，E AND G.W. Cσπ 阻 LL (eds ふ 1986. Check- Li st 01 birds 01 the world: A Continuation 01 the Work 01 James L Peters. Volume XI. Museum of Comparative Zoo logy ，Cambridge ，Massachusetts ，U.S. A. 0 四 N， D. F. 1963. 百 e rufous and white forms of an Asiatic Paradise F1 ycatcher ， Terpsiphone paradisi. Ardea 51: 51: 230 ー236. PERRINS ， C. M. 1965. Population fluctuations and clutch size in the Gre at Tit Parus majo れよAnim. Eco l. 34: 601 -64 7. RILEY ，J. H. 1938. Birds from Siam and the M a1 ay peninsula in 出巴 U 凶ted States National Museum collected by by Drs. Hugh M. Smith and William L. Abbott. Bull. U. ぷ Nat. Mus. 172. ROUND ， P. D. AND U. TREEs UCON. 1992. Th ung Tieo Nature Trai l. Kh ao Pr a Bang 悶 lram Non-Hunting Are a. 阻 ao Nor Ch uchi Lo wland Forest Pro ject ，τ 'h ailand. SALOMONSEN ， F. 1933. Revision of the group Tchitrea affinis Bly 白. Ibis: 730 ー745.