Pacific Science, vol. 69, no. 1 July 24, 2014 (Early view)  

1

Rapid change in song structure in introduced Japanese Bush-warblers (Cettia diphone) in

Hawaii1

By Shoji Hamao

Abstract Song birds that have been artificially introduced to isolated areas are fruitful material for investigating the changeability of songs within a limited period of time. I studied the songs of Japanese Bush-warblers, Cettia diphone, which were introduced from Japan to the island of Oahu (Hawain Islands) ca. 80 years ago . The warblers on Oahu sang acoustically simpler songs at lower frequencies than the warblers on Honshu, the main island of Japan. Previous studies found similar tendencies on small peripheral Japanese islands. Morphological characteristics indicated that the warblers on the Hawaiian Islands did not originate from insular subspecies in Japan. Therefore, the acoustic structure of their songs may have changed during their 80 years on Oahu. Possible factors driving this rapid change are relaxed sexual selection and/or the sound transmission properties of the island. Introduction

The acoustic structure of bird songs evolves under various selection pressures from ecological

and social factors (Kroodsma and Miller 1996). Because song structure is a culturally transmitted

trait in oscine passerines (Beecher and Brenowitz 2005), it can change rapidly relative to genetic

1Manuscript accepted 9 May 2014.

2

traits (Wilkins et al. 2013). Song birds that have been artificially introduced to isolated areas

(e.g., islands) are fruitful material for investigating the changeability of songs within a limited

period of time (Baker and Jenkins 1987). For example, the New Zealand population of

chaffinches, Fringilla coelebs, was introduced from England about 100 years ago, and their

songs have fewer trills and conclude with a more elaborate structure. This indicates that their

songs changed within approximately 100 years, likely due to the sound transmission properties

of the dense pine forests in New Zealand (Jenkins and Baker 1984).

The Japanese Bush-warbler, Cettia diphone, was introduced to the island of Oahu

(Hawaiian Islands) from Japan multiple times between 1929 and 1933 (Caum 1933), and it

currently inhabits many of the Hawaiian Islands (Pratt et al. 1987, Nelson and Vits 1998). In this

study, I examined whether the acoustic structure of songs differs between Oahu and Japan. To

assess the origin of the warblers in Hawaii, I also compared their morphology with that of

Japanese Bush-warblers in Japan. Finally, I discuss the possibility of song structure having

changed in Hawaii since their introduction and the factors responsible for such changes. I

predicted that song structure would differ between the Hawaiian and Japanese populations due to

rapid changes in songs in Hawaii during the ca. 80 years since colonization.

Materials and Methods

Field work

I conducted the study in Hawaii and Japan. Four study sites on Oahu were used in Hawaii:

Honolulu Watershed Forest Reserve (21°20'N, 157°50'W), Pupukea Forest Reserve (21°38'N,

158°01'W), Marconi Road (21°42'N, 157°59'W), and Mokuleia Forest Reserve (21°32'N,

158°11'W). At these sites, broad-leaved evergreen forests were the dominant vegetation. The

study site in Japan was located at Ranzan (36°01' – 06'N, 139°18' – 20'E), in central Honshu, the

3

largest island of Japan, where deciduous forests and bamboo thickets are the dominant

vegetation.

I collected song recordings from 24 males on Oahu during the breeding season (March 19

– 22) in 2010, and from 38 males at Ranzan during the breeding season (May 26 – June 16) in

2003. The songs sung by a male were recorded for at least 5 min. Each recording was sufficient

to obtain all song types of the male, as the repertoires of this species were small (2 – 5 song

types; Momose 1999), and the males sang very frequently (interval between songs: 14.7 s;

Hamao and Ueda 2000) and tended to avoid repetition of the same song types (Momose 1999). I

used a Sony (Tokyo, Japan) PCM-D50 digital recorder equipped with a Sony ECM-G5M

directional microphone on Oahu, and a Sony TCD-D7 DAT-recorder with a Sony ECM-G3M

directional microphone at Ranzan. To avoid duplicate sampling of the same male, I walked along

roads and recorded new birds when I encountered them. I used each route only once. Because

male territories were often contiguous, and males sang very frequently, I often heard the songs of

two or more males at the same time and therefore could identify individuals. Moreover, 84% of

males have a unique song type repertoire, which is useful for individual identification (Hamao

1993). Therefore, I could confirm whether a male was a new bird.

For body measurements, I captured 16 warblers using mist-nets at Ranzan during the

breeding seasons (end of April to end of July) between 2002 and 2004. I measured the lengths of

the wing (wing unflattened), tail, tarsus, and bill (exposed culmen) to the nearest 0.01 mm using

calipers. To obtain measurements of warblers in Hawaii, I used 15 specimens that were

preserved in the Bishop Museum, Honolulu. The measurements were made on March 20, 2010,

of specimens that had been collected from Oahu (1 male), Molokai (3), Hawaii (4), Maui (4), and

Kauai Islands (3) in 1992 (1 male), 2004 (7) and 2005 (7). The lengths of the wing, tail, and

tarsus were measured to the nearest 0.1 mm using dividers, and bill length was measured to the

4

nearest 0.01 mm using calipers. I obtained measurements only from males because capturing

females proved difficult and the museum contained few female specimens.

Song analyses

All recorded songs were analyzed using the Avisoft SASLab Pro software (ver. 5.1, Avisoft

Bioacoustics, Berlin, Germany). Sounds were displayed as sound spectrograms. Fast Fourier

transform (FFT) lengths of 256 and 1,024 were used to produce plots for temporal and frequency

measurements, respectively. The songs of Japanese Bush-warblers consist of an initial constant

frequency (CF) and subsequent frequency modulation (FM) portions (Hamao 1993, Momose

1999). The warblers have two basic song types: a type-H song and a type-L song. Type-H songs

contain one long whistle in the CF part, whereas type-L songs contain several short whistles

(Hamao 1993, Momose 1999; Figure 1). In Japan, individual males sing two to five different

song types that include at least one type-H and one type-L song, respectively (Momose 1999). I

recorded the number of song types for each male and quantified five acoustic variables in the FM

part of each song type. Maximum and minimum frequencies (Hz) were the highest and lowest

frequencies, respectively. I also measured the number of notes, the number of frequency

inflections [i.e., the number of changes in the sign of the derivative (slope) of the frequency on

the sound spectrogram], and song length (duration; Figure 1a). A note is a discrete song element,

and a continuous trace on a spectrogram (Catchpole 1980, Baker 1996).

Statistics

To compare song structure between Oahu and Ranzan, I used generalized linear mixed models

(GLMMs). The maximum and minimum frequencies, and the length of the FM parts were

analyzed using a GLMM with an identity link and a Gaussian error distribution. The number of

5

notes and frequency inflections were analyzed using a GLMM with a log link and a Poisson error

distribution. I included site (Oahu or Ranzan) as a fixed factor and male identity as a random

factor. To compare the number of song types among males, I used a generalized linear model

(GLM) with a log link and a Poisson error distribution. For these analyses, I created separate

models for type-H and type-L songs. To test the significance of any effects, I used likelihood

ratio tests.

To compare morphological measurements between Hawaii and Ranzan, I used

independent samples t-tests. In all analyses, the significance level was set at 0.05, and all

analyses were conducted using R software (ver. 3.0.1, R Core Team 2013).

Results

Songs sung by individual males on Oahu and at Ranzan included at least one type-H and one

type-L song, respectively (Figure 1, Supplemental Material: Media files S1 and S2). All five FM

part acoustic variables differed significantly between songs sung on Oahu and those sung at

Ranzan for both type-H and type-L songs (all P < 0.01). The maximum and minimum

frequencies were lower on Oahu than at Ranzan (Figure 2a, b). The length of the FM parts was

shorter on Oahu (Figure 2c). Both the numbers of notes and inflections were smaller on Oahu

than at Ranzan (Figure 2d, e). Thus, songs on Oahu had a lower frequency and a simpler acoustic

structure. The song type repertoire was larger on Oahu than at Ranzan for type-H songs (χ21 =

8.03, P < 0.01), but there was no difference between the sites for type-L songs (χ21 = 1.51, P =

0.22; Figure 2f).

In the body measurements of male warblers, wing, tail, and bill lengths were shorter in

the Hawaiian Islands than at Ranzan, whereas tarsus length was longer in the Hawaiian Islands

(Table 1). The plumage color of males at Ranzan was olive brown or slightly dark brown on the

6

upperparts and dirty white on the underparts. Hawaiian males had similar coloration, but they

were slightly more grayish on the upperparts.

Discussion

Japanese Bush-warblers on Oahu sang acoustically simpler songs at a lower frequency than the

warblers on Honshu, the main island of Japan (Figure 2). Previous studies using similar analyses

showed that the songs of warblers on small peripheral Japanese islands also had a simpler

structure compared with the songs sung on Honshu (Izu Islands, Ryukyu Archipelago: Hamao

2013, Ogasawara Islands: Hamao and Ueda 2000). Among these populations, birds on Honshu

and the Izu Islands belong to the subspecies cantans, and birds on the Ryukyu and Ogasawara

Islands belong to restricta and diphone, respectively (OSJ 2012).

The Japanese Bush-warbler includes six subspecies (OSJ 2012), but the subspecies or

source population of the warblers that were introduced to Hawaii was not recorded. However, the

plumage color and morphology of the subspecies provide some information about their identity.

Of the six subspecies, I excluded subspecies riukiuensis and borealis as candidates because the

former is thought to be a synonym of other subspecies (sakhalinensis and/or cantans: Kajita et al.

2002, Brazil 2009) and the latter is an accidental visitor to Japan (OSJ 2012). Subspecies

diphone differs from the birds in Hawaii in plumage color; it has a rufous-brown forehead and

crown, and a yellowish white underpart (Yamashina 1941, Brazil 2009). Also, diphone is much

smaller than the warblers in Hawaii (wing length of male diphone: 56.6 – 59.9 mm, n = 5,

Hamao unpubl. data). Subspecies restricta is similar to the warblers in Hawaii in body size (wing

length of male restricta: 60.6 – 66.7 mm, n = 36, Hamao unpubl. data). However, the plumage

color differs; the forehead, wing, and tail of this subspecies have a strong red-brown tone

(Yamashina 1941, Brazil 2009). Although the plumage colors of older (>50 years) museum

7

specimens are degraded, especially in UV wavelengths (McNett and Marchetti 2005, Armenta et

al. 2008), I was able to compare the colors of live birds and museum specimens in this study

because the specimens were collected recently (5 – 6 years ago, with one exception of 18 years

ago; see Materials and Methods) and because the differences in coloration between subspecies

were very clear. Thus, the Japanese Bush-warblers that were introduced to Hawaii appear to

belong to the subspecies sakhalinensis and/or cantans. The body size and plumage color of the

birds in Hawaii are consistent with those subspecies, although statistical analyses showed that the

wing, tail, and bill lengths were shorter, and tarsus length was longer in Hawaii than observed in

cantans in Japan (Table 1). These small differences may have been caused by shrinkages of

specimens (Winker 1993) or by the morphological changes as reported in Laysan finches,

Telespyza cantans (Conant 1988). DNA analyses will help to identify the origin of the Japanese

Bush-warblers that were introduced to Hawaii, but differences in genetic structure have not been

found among populations of cantans and sakhalinensis in analyses using the mitochondrial cyt b

region (Kajita 2002) or the CO1 region (Emura et al. 2013).

The songs sung in some insular populations of cantans have an acoustically simpler

structure than songs in the mainland population. Songs on Miyake-jima and Nii-jima Islands

consist of shorter FM parts and contain notes with fewer inflections than songs on Honshu

(Hamao 2013), which is similar to the tendency observed in the songs on Oahu (Figure 2). These

islands are located 150 – 180 km south of Tokyo. Access to these islands would not have been

easy at the time when the warblers were introduced to Hawaii (ca. 80 years ago). Japanese Bush-

warblers were systematically introduced to Oahu multiple times by the Board of Agriculture and

Forestry and Hui Manu, a group that imported song birds (Caum 1933). It is reasonable to

conclude that the warblers on Oahu originated from populations on the mainland of Japan, and

8

that their songs have changed during the past 80 years, although I cannot rule out the possibility

that the Hawaiian warblers were brought from small peripheral Japanese islands.

A factor related to the simplified songs on Oahu may be the residency of the birds. In

mainland Japan, Japanese Bush-warblers move seasonally and males establish breeding

territories every spring (Hamao 1992, Yamashina Institute for Ornithology 2002). Moreover,

territorial males are frequently replaced (Hamao 1992). Males are exposed to strong male-male

competition. In contrast, competition may be weaker among males on Oahu because they do not

migrate and their density is relatively low (S. H., pers. obs.). As complex songs indicate high-

quality males (Behr et al. 2006, Cardoso et al. 2012) and are therefore considered advantageous

in male–male competition, the simplified songs on Oahu may be the result of relaxed male–male

competition. In Japan, insular populations of Japanese Bush-warblers show a small degree of

sexual size dimorphism, suggesting that sexual selection is relaxed on the islands (Hamao 2013).

Studies of the territoriality and sexual dimorphism in the Hawaiian population are required.

Song structure is also shaped by the sound transmission properties of the habitat (Jenkins

and Baker 1984, Kroodsma 1985). Birds inhabiting forests with dense foliage use pure tone-like,

low-frequency sounds that attenuate less in their habitats (Morton 1975, Richard and Wiley

1980). The warblers inhabit broad-leaved evergreen forests on Oahu, but deciduous forests and

bamboo thickets on the mainland of Japan. It is possible that the vegetation structure on Oahu

affects sound transmission and has driven the change to low-frequency, less modulated songs.

More information is required about vegetation structure and sound transmission in the habitats

on Oahu and in Japan.

Characteristics of introduced populations are often restricted because of the small number

of colonizing individuals (founder effect). For example, the chaffinches that colonized the

Chatham Islands in about 1900 have a smaller number of syllable types than the source

9

population in New Zealand (Baker and Jenkins 1987). Although the song type repertoire has not

been investigated at the population level in Japanese Bush-warblers, individual males on Oahu

had a larger number of song types than males in Japan (Figure 2f). It is doubtful that this pattern

can be explained by the founder effect.

It is possible that the observed changes were the result of random cultural drift (Baker et

al. 2001, Benedict and Bowie 2009). However, I consider this unlikely, as multiple acoustic

variables, including shorter duration, fewer notes, and less modulation, indicated a simplified

song structure.

Cultural evolution of songs can occur rapidly. The acoustic structure of the songs of

white-crowned sparrows, Zonotrichia leucophrys oriantha, changed within 26 years in California

(Harbison et al. 1999). The white-crowned sparrows, Z. l. pugetensis, in Oregon and Washington

had six song dialects but showed little genetic differentiation in four microsatellite loci (Soha et

al. 2004). Therefore, it is likely that, during the 80 years since colonization, the songs of

Japanese Bush-warblers on Oahu have undergone acoustic changes.

Acknowledgments

I thank Lydia Garetano for allowing me to measure specimens in the Bishop Museum, and Jeff

Foster for providing me a lot of useful information and literatures. I am indebted to Eric

VanderWerf and Hiroshi Uchida for their support during my field work, and two anonymous

reviewers for their constructive comments. This study was supported by funds from the

Yamashina Foundation for Wildlife Conservation Research.

References

10

Armenta, J. K., P. O. Dunn, and L. A. Whittingham. 2008. Effects of specimen age on plumage

color. Auk 125:803 – 808.

Baker, A. J., and P. R. Jenkins. 1987. Founder effect and cultural evolution of songs in an

isolated population of chaffinches, Fringilla coelebs, in the Chatham Islands. Anim.

Behav. 35:1793 – 1803.

Baker, M. C. 1996. Depauperate meme pool of vocal signals in an island population of singing

honeyeaters. Anim. Behav. 51:853 – 858.

Baker, M. C., E. M. Baker, and M. S. A. Baker. 2001. Island and island-like effects on vocal

repertoire of singing honeyeaters. Anim. Behav. 62:767 – 774.

Benedict, L., and R. C. K. Bowie. 2009. Macrogeographical variation in the song of a widely

distributed African warbler. Biol. Lett. 5:484 – 487.Brazil, M. 2009. Birds of East Asia.

Christpher Helm, London.

Beecher, M. D., and E. A. Brenowitz. 2005. Functional aspects of song learning in songbirds.

Trends Ecol. Evol. 20:143 – 149.

Behr, O., O. von Helversen, G. Heckel, M. Nagy, C. C. Voigt, and F. Mayer. 2006, Territorial

songs indicate male quality in the sac-winged bat Saccopteryx bilineata (Chiroptera,

Emballonuridae). Behav. Ecol. 17:810 – 817.

Cardoso, G. C., J. W. Atwell, Y. Hu, E. D. Ketterson, and T. D. Price. 2012. No correlation

between three selected trade-offs in birdsong performance and male quality for a species

with song repertoires. Ethology 118:584 – 593.

Catchpole, C. K. 1980. Sexual selection and the evolution of complex songs among European

warblers of the genus Acrocephalus. Behaviour 74:149 – 165.

Caum, E. L. 1933. The exotic birds of Hawaii. Bernice P. Bishop Mus. Occ. Pap. 10:1 – 55.

11

Conant, S. 1988. Geographic variation in the Laysan finch (Telespyza cantans). Evol. Ecol.

2:270 – 282.

Emura, N., H. Ando, K. Kawakami, and Y. Isagi. 2013. Genetic and morphological differences

among populations of the Japanese Bush-warbler (Aves: Sylviidae) on the Ogasawara

Islands, Northern Pacific. Pacific Science 67:187 – 196.

Hamao, S. 1992. Lack of pair-bond: a polygynous mating system of the Japanese bush warbler

Cettia diphone. Jpn. J. Ornithol. 40:51 – 65.

Hamao, S. 1993. Individual identification of male Japanese bush warbler Cettia diphone by song.

Jpn. J. Ornithol. 41:1 – 7.

Hamao, S. 2013. Acoustic structure of songs in island populations of the Japanese bush warbler,

Cettia diphone, in relation to sexual selection. J. Ethol. 31:9 – 15.

Hamao, S., and K. Ueda. 2000. Simplified song in an island population of the bush warbler

Cettia diphone. J. Ethol. 18:53 – 57.

Harbison, H., D. A. Nelson, and T. P. Hahn. 1999. Long-term persistence of song dialects in the

mountain white-crowned sparrow. Condor 101:133 – 148.

Jenkins, P. F., and A. J. Baker. 1984. Mechanisms of song differentiation in introduced

populations of chaffinches Fringilla coelebs in New Zealand. Ibis 126:510 – 524.

Kajita, M. 2002. Intra-specific phylogeography in Cettia diphone. Iden 56:42 – 46.

Kajita, M., T. Mano, and F. Sato. 2002. Two forms of bush warbler Cettia diphone occur on

Okinawajima Island: re-evaluation of C. d. riukiuensis and C. d. restricta by multivariate

analyses. J. Yamashina Inst. Ornithol. 33:148 – 167.

Kroodsma, D. E. 1985. Geographic variation in songs of the Bewick’s wren: a search for

correlations with avifaunal complexity. Behav. Ecol. Sociobiol. 16:143 – 150.

12

Kroodsma, D. E., and E. H. Miller (eds). 1996. Ecology and evolution of acoustic

communication in birds. Cornell University Press, Ithaca.

McNett, G. D., and K. Marchetti. 2005. Ultraviolet degradation in carotenoid patches: live versus

museum specimens of Wood Warblers (Parulidae). Auk 122:793 – 802.

Momose, H. 1999. Structure of territorial songs in the Japanese bush warbler (Cettia diphone).

Mem. Fac. Sci. Kyoto Univ. (Ser. Biol.) 16:55 – 65.

Morton, E. S. 1975. Ecological sources of selection on avian sounds. Am. Nat. 109:17 – 34.

Nelson, J. T., and A. Vitz. 1998. First reported sighting of Japanese bush-warbler (Cettia

diphone) on the island of Hawaii. 'Elepaio 58:1 – 2.

OSJ (The Ornithological Society of Japan). 2012. Check-list of Japanese birds, 7th revised

edition. The Ornithological Society of Japan, Sanda.

Pratt, H. D., P. L. Bruner, and D. G. Berrett. 1987. A field guide to the birds of Hawaii and the

tropical Pacific. Princeton University Press, Princeton, New Jersey.

R Core Team. 2013. R: A language and environment for statistical computing. R Foundation for

Statistical Computing, Vienna, Austria. URL http://www.R-project.org/.

Richard, D. G., and R. H. Wiley. 1980. Reverberations and amplitude fluctuations in the

propagation of sound in a forest: implications for animal communication. Am. Nat.

115:381 – 399.

Soha, J. A., D. A. Nelson, and P. G. Parker. 2004. Genetic analysis of song dialect populations in

Puget Sound white-crowned sparrows. Behav. Ecol. 15:636 – 646.

Wilkins, M. R., N. Seddon, and R. J. Safran. 2013. Evolutionary divergence in acoustic signals:

causes and consequences. Trends Ecol. Evol. 28:156 – 166.

Winker, K. 1993. Specimen shrinkage in Tennessee Warblers and “Traill’s” flycatchers. J. Field

Ornithol. 64:331 – 336.

13

Yamashina, Y. 1941. A natural history of Japanese birds. Iwanamishoten, Tokyo.

Yamashina Institute for Ornithology. 2002. Atlas of Japanese migratory birds from 1961 to 1995.

Yamashina Institute for Ornithology, Abiko.

Pacific Science, vol. 69, no. 1 July 24, 2014 (Early view)  

14

TABLE 1 Differences in body measurements of male Japanese Bush-warblers between the Hawaiian Islands and Ranzan on the Japanese mainland. Means and ranges (mm) are shown for each variable. Numbers in parentheses are the numbers of individuals. Asterisks indicate significant differences detected by t-tests: ** P < 0.01, *** P < 0.001.

Tarsus Wing1 Tail Bill2

Hawaiian Islands 25.8 61.3 62.2 10.9 (15)3 24.7 – 27.9 59.4 – 64.0 55.7 – 67.2 10.1 – 11.4 Ranzan 25.0** 63.3*** 68.0*** 11.6***

(16)4 23.7 – 25.9 59.7 – 65.4 65.4 – 71.8 10.9 – 12.4

1 Wing unflattened.. 2 Exposed culmen. 3 Measurements of tarsus and culmen were obtained from 14 males, because the specimens were damaged. 4 Measurements of tail were obtained from 14 males, because two males were molting their tails.

15

Figure 1. Spectrograms of Japanese Bush-warbler songs showing two basic song types

and the parameters analyzed. (a) A type-H song and (b) a type-L song from a male on

the island of Oahu. The type-H song contains two notes and two inflections in its

frequency modulated (FM) portion. See the supplemental material that includes the

sounds of these songs.

16

Figure 2. Differences in song structure and repertoire size in Japanese Bush-warblers

between the island of Oahu and Japan: maximum frequency (a), minimum frequency

(b), song length (duration; c), number of notes (d), number of inflections (e), and

number of song types (f). The acoustic variables (a-e) were obtained in FM portions of

songs. Means +- 1SD are shown. White and gray bars indicate type-H and type-L songs,

respectively. Horizontal solid and dotted lines indicate significant differences in type-H

and -L songs, respectively (** P < 0.01, *** P < 0.001).

17

Supplementary material associated with this article is available.

S1 (MS1213_spl1.mp3)

A type-H song from a Japanese Bush-warbler on the island of Oahu. This sound

corresponds with the song in Figure 1a.

S2 (MS1213_spl2.mp3)

A type-L song from a Japanese Bush-warbler that is the same male in S1. This sound

corresponds with the song in Figure 1b.