Skip to main content
NCBI home page
Zoological Studies logoLink to Zoological Studies
. 2016 Aug 22;55:e38. doi: 10.6620/ZS.2016.55-38

Song Characteristics of Oriental Cuckoo Cuculus optatus and Himalayan Cuckoo Cuculus saturatus and Implications for Distribution and Taxonomy

Canwei Xia 1, Wei Liang 2, Geoff J Carey 3, Yanyun Zhang 1,*
PMCID: PMC6511894  PMID: 31966183

Abstract

Canwei Xia, Wei Liang, Geoff J. Carey, and Yanyun Zhang (2016) Song features during the breeding season are important in identifying species of cuckoos. Whether Oriental Cuckoo Cuculus optatus and Himalayan Cuckoo C. saturatus inhabiting the Palearctic and Oriental realms respectively can be distinguished according to song characteristics is uncertain. In this study, we performed a thorough investigation of the song characteristics of these taxa by collecting and analyzing recordings of song in their distribution areas. We found that songs could be divided into two groups based on the number of notes per syllable, and significant differences in other frequency and temporal features were also found between these two groups. The group with a song comprising two notes per syllable was shown to breed in Russia, Japan and China including Xinjiang, Inner Mongolia, Heilongjiang and Taiwan, while the group with a song containing more than two notes per syllable was found to breed in the Himalayas and central China, extending northeast through north China as far as northeast Hebei, and south to southwest China. The distribution of these two groups was broadly related to the published distribution of populations of optatus and saturatus, respectively. Our data supported the separation of optatus and saturatus based on their song features, and also suggested refinements to the distribution of these two taxa, as follows: birds in north mainland China are saturatus, and those in Taiwan Island are optatus.

Keywords: Distribution, Himalayan cuckoo, Oriental cuckoo, Song characteristics, Taxonomy

BACKGROUND

Song is a vocal behavior of birds that serves as a means of territory defense and/or mate attraction during the breeding season (Marler and Slabbekoorn 2004; Catchpole and Slater 2008). As an important reproductive isolation mechanism, song is an essential basis for classification among Passeriformes (Slabbekoorn and Smith 2002; Alström and Ranft 2003). In non-Passeriformes, vocalizations are often simple and stereotyped, with no song behavior. However, certain non- Passeriformes, such as waders, rails, owls, pigeons and barbets, utilize a song that, although stereotyped, is important in territory defense (Kroodsma and Miller 1996). This is the case with cuckoos (Cuculus spp.), whose vocalizations during the breeding season serve as a means of territory defense and mate attraction, similar to the songs of Passeriformes (Payne 2005). Since cuckoo song is relatively conserved with little intraspecies variation, it can serve as an important basis for interspecies classification (Payne 2005).

Oriental Cuckoo Cuculus optatus (previously C. horsfieldi, a junior synonym) and Himalayan Cuckoo Cuculus saturatus were originally described as species before Peters (1940) treated optatus as a subspecies of saturatus. The latter approach was generally followed into the 1990s (e.g. Sibley and Monroe 1990; Howard and Moore 1991) until Payne (1997) accorded optatus specific status based on differences in song. Johnsgard (1997) classified the taxa according to their distributions during the breeding season, with optatus breeding in the Palearctic realm and saturatus in the Oriental realm. King (2005) and Payne (2005) provided further evidence of their specific status based on song features, and this is the approach of the International Ornithological Congress (Gill and Donsker 2014), which has adopted Oriental Cuckoo (optatus) and Himalayan Cuckoo (saturatus) as the English names, which we follow here. However, Lindholm and Lindén (2007) found that song of presumed optatus in Hebei and Shanxi, mainland China (located in the Palearctic realm) showed greater resemblance to the song of saturatus, whereas the song of presumed saturatus from Taiwan Island (located in the Oriental realm) resembled more that of optatus, implying song features are highly variable within species. Based on this research, Erritzøe et al. (2012) believed there was an overlap between the songs of saturatus and optatus, and treated optatus as a subspecies of saturatus. As suggested by Payne (2005), in cuckoos, “if populations have the same calls and songs, they are considered one species, and if they have different calls, they are different species.” Thus, whether there are consistent differences in song between saturatus and optatus is of importance in determining whether they should be treated as one or two species.

Cuculus optatus and C. saturatus have significant differences neither in mitochondrial DNA (Payne 2005) nor in morphology, other than the larger size of the former (Leven 1998); as a result, song differences between the two have been the focus of taxonomic discussions (King 2005; Payne 2005; Erritzøe et al. 2012). Payne (2005) classified optatus and saturatus as two species based on the following (see Fig. 1 for song terms): (1) the first note in syllable of saturatus has a higher frequency than the second note, which is not observed in optatus; and (2) the syllable frequency of optatus is lower than that of saturatus. King (2005) also supported this distinction between the two species, and believed that song differences could be summarized as follows: (1) the first note of each syllable of saturatus has a higher frequency than the second, which is not observed in optatus, and (2) the number of notes per syllable of optatus is strictly two, while that of saturatus is greater than two. In this study, we aimed to perform a thorough analysis of the song of saturatus and optatus in order to shed further light on distribution and classification. We tested whether songs can be classified into two groups based on the three features proposed by Payne (2005) and King (2005): (1) the number of notes per syllable; (2) the frequency ratio of the first note over the second note; and (3) the syllable frequency. On the basis of above research, we also analyzed other differences in song features and discussed distribution areas and taxonomy of these two groups accordingly.

Fig. 1.

Fig. 1.

Open in a new tab

Fig. 1. Spectrograms of the songs of Cuculus optatus and Cuculus saturatus and song terms. The red dashed line indicates the peak frequency of note.

MATERIALS AND METHODS

Audio acquisition

We obtained 119 recordings from our own data as well as online databases (Macaulay Library [http://macaulaylibrary.org], Xeno-canto [http://www.xeno-canto.org], and AVoCet: Avian Vocalizations Center [http://avocet.zoology. msu.edu]). Goldwave 5.25 audio processing software (Goldwave, Canada) was used for audio resampling with the following parameters: sampling precision, 16 bit; sampling frequency, 8,000 Hz; and audio format, .wav. For multiple recordings collected on the same day at the same location such that individuals could not be identified, one recording was randomly selected for analysis to avoid false repetition. Recordings with background noise that affected measuring were also removed. A total of 83 usable recordings were obtained (Additional file 1).

As differences between optatus and saturatus in molecular date (e.g. mitochondrial DNA, Payne 2005) is not significant, in plumage and size are minor, and song features are overlap (Lindholm and Lindén 2007; Erritzøe et al. 2012), now, the taxa are mainly based on distribution areas. One aim of this study is to refine distribution areas, so we didn’t link the recordings with the corresponding species based on names in online databases, which is mainly based on distribution areas. We firstly mixed all recordings together, and then linked the recordings with the corresponding taxa based on song features, if there are any song features can successfully split the group.

Audio measurements

Avisoft-SASLab Pro 5.2 audio analysis software (Avisoft Bioacoustics, Germany) was used to generate spectrograms. Different spectrogram parameters were used to measure temporal and frequency features to achieve higher precision. The spectrogram parameters for temporal feature measurement were as follows: Hamming window; FFT length, 64; frequency bandwidth, 163 Hz; and time resolution, 4 ms. The spectrogram parameters for frequency feature measurement were as follows: Hamming window; FFT length, 1,024; frequency bandwidth, 10 Hz; and time resolution, 64 ms. Terms used in analysis are illustrated in figure 1. A note refers to a trace continuously appearing in the spectrogram, while a syllable is composed of multiple notes, with relatively longer time intervals between syllables. Based on previous studies (King 2005; Lindholm and Lindén 2007) and the requirements of this study, an automatic measurement method was applied to measure the following variables: duration and peak frequency of the first note; duration and peak frequency of the second note; and duration, peak frequency and note number of syllable. Peak frequency refers to the frequency associated with the maximum energy, and be called frequency for short below. The frequency ratio of the first note relative to the second note was also calculated.

As noted by Lindholm and Lindén (2003), the song of ‘Oriental Cuckoo’ sensu lato usually begins with the ‘soft phrase’ (a softer and faster sequence of 4-10 notes), followed by a series of syllables. The soft phrase is uttered once only in each song series, and, given the unpredictable nature of cuckoo song utterance, was consequently absent in many of our recordings. Our pragmatic solution to this problem was to focus on song features in the remainder of the song, an approach also adopted by King (2005) and Lindholm and Lindén (2003, 2007).

A sample of 39 individuals that had more than five consecutive syllables clearly recorded was first selected, and five consecutive syllables were measured for each individual. After demonstrating that song variables were highly conserved in the same recording (Table 1) using reliability analysis (Rankin and Stokes 1998), one syllable was randomly selected and measured from each of the recordings (sample size 83).

Table 1. Reliability of song features within recording.

Reliability Frequency of the first note Duration of the first note Frequency of the second note Duration of the second note Note number per syllable Sentence frequency Sentence duration First note frequency/ second note frequency
Mean 0.980 0.870 0.991 0.851 0.971 0.988 0.974 0.960
95% low 0.968 0.805 0.986 0.778 0.954 0.981 0.959 0.937
95% up 0.988 0.922 0.995 0.910 0.983 0.993 0.985 0.976
Open in a new tab

Data analysis

The following frequency distribution was proposed for the selected syllables from recordings: (1) the number of notes per syllable; (2) the frequency ratio of the first note relative to the second note in each syllable; and (3) the syllable frequency. Song features were plotted based on the quartiles of the above-mentioned variables.

Note number per syllable was found to have a bimodal distribution; accordingly, vocalizations were classified into two groups based on this, and differences in other song features were compared between these two groups. We applied a MANOVA to assess the overall differences between the two groups, followed by an independent sample t-test for each variable. Discriminant and principal component analyses were employed to examine whether features other than note number per syllable differed between the two groups. In discriminant analysis, results from jack-knifed classifications, in which each recording was assigned to a group on the basis of a discriminant function calculated from all recordings in the data set except the one being classified, are reported as percentages of recordings correctly assigned. In principal component analyses, varimax rotation was used, because it provided a clearer (very high loadings for some of the variables and very low loadings for the others) interpretation of the principal components (PCs) than without varimax rotation.

Statistical analysis was performed using SPSS 21.0. Data were presented as the mean ± standard deviation. Differences with P values of less than 0.05 were considered significant. Permission for this study (recording cuckoo vocalizations) was granted by Ethic and Animal Welfare Committee in Beijing Normal University (license number CLS-EAW-2013-016).

RESULTS

Note number per syllable was higher in Himalayan areas (where 4-6 notes were recorded) and central China (where four notes dominated), and lower in Russia, Japan and some region of China including Xinjiang, Inner Mongolia, Heilongjiang and Taiwan, where only two notes per syllable were recorded (Fig. 2). The frequency ratios of the first note relative to the second note were higher in Himalayan areas, central China and Beijing/Hebei, and lower in Russia, Japan and some region of China including Xinjiang, Inner Mongolia, Heilongjiang and Taiwan, respectively (Fig. 3). The highest syllable frequency was observed in Taiwan Island, while the lowest syllable frequency was observed in Russia; however, syllable frequency varied substantially among adjacent locations in Himalayan areas and central China (Fig. 4). There was a significant correlation between the note number per syllable and the frequency ratio of the first note relative to the second note (Kendall correlation coefficient = 0.612, P < 0.001). In contrast, there were no significant correlations of either of these parameters with syllable frequency (Kendall correlation coefficient = 0.079, P = 0.317; Kendall correlation coefficient = 0.081, P = 0.360, respectively).

Fig. 2.

Fig. 2.

Open in a new tab

Fig. 2. Frequency distribution (A) and geographical distribution (B) of the note number per syllable in each syllable. The red arrow indicates the quartile. Geographic base map in (B) is from Google Maps (Google, USA).

Fig. 3.

Fig. 3.

Open in a new tab

Fig. 3. Frequency distribution (A) and geographical distribution (B) of the frequency ratio of the first note over the second note. The red arrow indicates the quartile. Geographic base map in (B) is from Google Maps (Google, USA).

Fig. 4.

Fig. 4.

Open in a new tab

Fig. 4. Frequency distribution (A) and geographical distribution (B) of syllable frequency. The red arrow indicates the quartile. Geographic base map in (B) is from Google Maps (Google, USA).

Among the aforementioned variables (Figs. 2A, 3A, and 4A), note number per syllable showed a clear bimodal distribution with peak values of 2 and 4. Furthermore, after mapping note number per syllable (Fig. 2B), it was clear that there was broad congruence with published distributions of optatus and saturatus, and with the differentiation described by King (2005), with birds having two notes per syllable thus considered referable to optatus, and birds with more than two per syllable referable to saturatus. Consequently, we set note numbers of two or more than two per syllable as criteria to divide birds into two groups (n = 42 and n = 41, respectively).

Other song features (except for the number of notes per syllable) are significantly different between two groups (MANOVA: Pillai’s Trace = 0.934, F7, 75 = 152.315, P < 0.001). For the group with two notes, the first note had a relatively lower frequency and longer duration, the second note had a relatively shorter duration, the duration of the syllable was relatively short, and the frequency ratio of the first note over the second note was relatively small (Table 2). Discriminant analysis using other song features (except for the number of notes per syllable) achieved 100% accuracy in distinguishing between the two groups. Principal components analysis generated three principal components with eigenvalues > 1.0, explaining 92.4% of the variance in the original seven variables (Table 3). The two groups could also be clearly separated by conducting principal component analysis with song features other than note number per syllable and plotting with the first two principal components (Fig. 5).

Table 2.  Comparison between song variables of different cuckoo groups (independent sample t-test) .

Song variables Group 1 (Mean ± SD) Group 2 (Mean ± SD) t81 P
Frequency of the first note (Hz) 428.56 ± 48.09 472.02 ± 32.39 -4.840 < 0.001
Duration of the first note (s) 0.08 ± 0.02 0.04 ± 0.01 11.329 < 0.001
Frequency of the second note (Hz) 426.29 ± 45.97 421.88 ± 21.06 0.564 0.574
Duration of the second note (s) 0.09 ± 0.02 0.11 ± 0.02 -3.292 0.001
Syllable frequency (Hz) 426.29 ± 45.97 422.26 ± 21.19 0.515 0.608
Syllable duration (s) 0.30 ± 0.03 0.82 ± 0.14 -23.621 < 0.001
First note frequency/second note frequency 1.00 ± 0.01 1.12 ± 0.06 -12.791 < 0.001
Open in a new tab

Table 3. Eigenvalues and factor loadings for principal components analysis of song variables.

Song variables PC1 PC2 PC3
Frequency of the first note (Hz) 0.828 0.501 0.171
Duration of the first note (s) -0.048 -0.919 0.189
Frequency of the second note (Hz) 0.993 -0.054 -0.055
Duration of the second note (s) -0.013 0.128 0.971
Syllable frequency (Hz) 0.994 -0.047 -0.043
Syllable duration (s) 0.003 0.893 0.193
First note frequency/second note frequency 0.075 0.851 0.336
Eigenvalue 3.155 2.309 1.007
% Variance explained 45.077 32.983 14.384
Open in a new tab

Fig. 5.

Fig. 5.

Open in a new tab

Fig. 5. Differentiation between different groups based on the first two principal components of song features (A) and geographical distribution (B). Geographic base map in (B) is from Google Maps (Google, USA). The relationship between principal components and song variables are shown in table 3.

DISCUSSION

Song difference between optatus and saturatus

In this study, we performed an in-depth investigation of song in optatus and saturatus. Our data showed that song could convincingly be used to distinguish birds into two groups, and that the distributions of these two groups was broadly related to the published distribution of populations of optatus and saturatus, respectively. Unlike the previous studies (e.g. King 2005; Lindholm and Lindén 2007), which have broadly accepted published distributions (Johnsgard 1997) and interpreted vocal data accordingly, we firstly tested whether song features could be used to classify recordings into two groups, and then refined the distribution areas of the two groups based on song features.

Cuckoo songs were quantified, and note number per syllable was found to have a clear bimodal distribution; thus, we used note number per syllable as the criterion for differentiation. Groups divided by this criterion also had significant differences in other song features, and the accuracy of discriminant analysis was as high as 100%. The method used to distinguish between saturatus and optatus, i.e., using note number per syllable, was the same in this study as in a previous study by King (2005). Also consistent with the latter, we found that the song of optatus had two notes, while the song of saturatus had more than two notes. Because it is quite straightforward to identify the number of notes, this criterion can also be used to distinguish singing birds in the field easily. The frequency ratio of the first note relative to the second note was also a criterion used by Payne (2005) and King (2005). According to data from our study, this feature was positively correlated with note number, and the distributions of the two parameters were generally consistent. Hence, the frequency ratio of the first note relative to the second note can also be used as a criterion to distinguish between the two groups. Payne (2005) stated that song frequency of optatus was lower than that of saturatus. According to our data, frequency tended to be higher in southern regions and lower in northern regions; however, substantial variations were found across adjacent areas, especially in Himalayan areas. For this reason, we cannot unequivocally recommend using song frequency to distinguish between these two groups.

Distributions and taxonomy of optatus and saturatus

Johnsgard (1997) classified optatus and saturatus as two subspecies according to their distributions during the breeding season, with optatus breeding in the Palearctic realm and saturatus in the Oriental realm; this division of the distribution areas has generally been utilized by other authors (e.g. Payne 2005; Erritzøe et al. 2012). Both King (2005) and Lindholm and Lindén (2007) found that the song of cuckoos in Taiwan Island (located in the Oriental realm), which had syllables composed of two notes, is more similar to optatus; while Lindholm and Lindén (2007) also found that the song of cuckoos in Hebei and Shanxi (located in the Palearctic realm), which had syllables composed of three to four notes, is similar to saturatus. Results of our study supported these findings, as all 11 recordings made in Taiwan Island had syllables composed of two notes, and six of seven recordings from Beijing and Hebei had syllables composed of three or four notes. The one recording that had syllables composed of two notes, recorded in May on a small island (‘Happy Island’ or Shi Jiu Tuo) off the coast of Hebei, is likely to have been a migrant rather than a locally-breeding bird, as the island appears to lack breeding habitat for both Oriental cuckoo and Himalayan cuckoo. Unlike Erritzøe et al. (2012) who believed that song is not discriminant between optatus and saturatus, we believe a simpler and more straightforward explanation is that birds in north mainland China are, indeed, saturatus, and those in Taiwan Island are optatus. We thus set the distribution boundary between optatus and saturatus in Northern China (Fig. 6), which is in line with distribution boundary of some species pairs based on recent researches, for example, Blue- and-white Flycatcher (Cyanoptila cyanomelana) and Zappey’s Flycatcher (Cyanoptila cumatilis) (Leader and Carey 2012), Songar Tit (Parus songarus) and Willow Tit (Parus montanus) (Kvist et al. 2001), Silver-throated Bushtit (Aegithalos glaucogularis) and Long-tailed Tit (Aegithalos caudatus) (Päckert et al. 2010). We obtained recordings from optatus and saturatus across most range of their breeding area, the exception being central Russia. This defect was alleviated by examining songs from 58 males in Russia presented in Lindholm and Lindén (2007). All these belong to optatus according the criteria in this study, as there are two notes per syllable and a smaller frequency change between the first note and the second note (Lindholm and Lindén 2007).

Fig. 6.

Fig. 6.

Open in a new tab

Fig. 6. Boundary in China between distribution areas of Cuculus optatus and Cuculus saturatus during breeding season. The dashed line is the boundary that Johnsgard (1997) used to separate Cuculus optatus and Cuculus saturatus. The solid line is the boundary suggested by this study. a: Northern mainland China, b: Taiwan Island. Geographic base map is from Google Maps (Google, USA).

As cuckoo song is innate and may have a strictly genetic basis (Mayr 1942), it is most useful in identifying species (Payne 2005). With regard to optatus and saturatus, Payne (1997) accorded both specific status mostly based on differences in their songs, while Erritzøe et al. (2012) treated the two as a single species after Lindholm and Lindén (2007) found apparent overlap between their songs. In this study, after a thorough investigation of the song characteristics of these taxa, we found the songs can clearly be split into two groups, and there is no song characteristic overlap between optatus and saturatus after a fine adjustment in distribution area of these two taxa. Thus, our data support the separation of optatus and saturatus into two species, as suggested by Payne (1997, 2005). Three notes per syllable are present in three recordings from Beijing and Hebei, near the northernmost part of the range of saturatus. The number of notes per syllable in these recordings lies between optatus (with two notes per syllable) and other saturatus (with four to six notes per syllable). This may indicate past interbreeding between optatus and saturatus at this area. Although song difference is not absolutely definitive in taxonomy without other evidence such as morphology and genetic markers, given the high degree of conformity of song with published distributions, there is strong evidence to support the conclusions. At least, our data deny there is an overlap between the songs of saturatus and optatus, which was proposed by Lindholm and Lindén (2007) and Erritzøe et al. (2012).

CONCLUSIONS

In this study, the song features of birds from across the range of optatus and saturatus were analyzed. Based on note number per syllable, birds could be split into two groups, and significant differences in other song features were also found between these two groups. The group with a song comprising two notes per syllable was shown to breed in Russia, Japan and China including Xinjiang, Inner Mongolia, Heilongjiang, Taiwan Island, while the group with a song containing more than two notes per syllable was found to breed in the Himalayas and central China, extending northeast through north China as far as northeast Hebei, and northern southeast Asia. The data largely agree with the size-related distribution described by Leven (1998), and with published distributions of optatus and saturatus, respectively (Johnsgard 1997; Payne 2005; Erritzøe et al. 2012). As the songs can be clearly split into two groups, our data support Payne (1997, 2005) in treating optatus and saturatus as two species.

Additional files

Additional file 1.

Recordings used in this study and original measurement data. (download)

zoolstud-55-038-s001.xls (39.5KB, xls)

Acknowledgments

Acknowledgment: The recordings from online databases (Macaulay Library [http://macau laylibrary.org], Xeno-canto [http://www.xeno- canto.org], and AVoCet: Avian Vocalizations Center [http://avocet.zoology.msu.edu]) used in this study were much appreciated. We thank Kuankuoshui and Dongzhai National Nature Reserves for support to carry out this study, Lijun Zhang and Chentao Wei for offering their recordings. Valuable comments on a draft were provided by Michael Leven, Paul Leader and Per Alström, and information was provided by Paul Holt. This work was funded by the National Natural Science Foundation of China (Nos. 31272328 and 31472013 to WL, 31172098 to YZ).

References

  1. Alström P, Ranft R. The use of sounds in avian systematics and the importance of bird sound archives. Bull Br Orn Club. 123:114–135. [Google Scholar]
  2. Catchpole C K, Slater Pjb. Bird song: biological themes and variations. Cambridge): Cambridge University Press; 2008. [Google Scholar]
  3. Erritzøe J, Mann C F, Brammer F P, Fuller R A. Cuckoos of the world. London): Christopher Helm; 2012. [Google Scholar]
  4. Gill F, Donsker D. IOC World Bird List (v 4.1). Available at http://www.worldbirdnames.org. Accessed 1 June 2015.
  5. Howard R, Moore A. A complete checklist of the birds of the world. London): Academic Press; 1991. [Google Scholar]
  6. Johnsgard P A. The avian brood parasites: deception at the nest. Oxford): Oxford University Press; 1997. [Google Scholar]
  7. King B. The taxonomic status of the three subspecies of Cuculus saturatus. Bull Br Orn Club. 125:48–55. [Google Scholar]
  8. Kroodsma D E, Miller E H(eds) Ecology and evolution of acoustic communication in birds. New York: Cornell University Press; 1996. [Google Scholar]
  9. Kvist L, Martens J, Ahola A, Orell M. Phylogeography of a Palaearctic sedentary passerine, the willow tit (Parus montanus) J Evolution Biol. 14:930–941. [Google Scholar]
  10. Leader P J, Carey G J. Zappey's Flycatcher Cyanoptila cumatilis, a forgotten Chinese breeding endemic. Forktail. 28:121–128. [Google Scholar]
  11. Leven M R. Asian lesser cuckoo: the first record for Hong Kong. pp. 118–124.
  12. Lindholm A, Lindén A. Oriental Cuckoo in Finland. Alula. 9:122–133. [Google Scholar]
  13. Lindholm A, Lindén A. Some notes on the distribution and songs of two Oriental Cuckoo taxa, Cuculus (saturatus) saturatus and Cuculus (saturatus) optatus. Forktail. 23:12–16. [Google Scholar]
  14. Marler P, Slabbekoorn H (eds) Nature's music: the science of birdsong. London): Elsevier Academic Press; 2004. [Google Scholar]
  15. Mayr E. Systematics and the origin of species. New York): Columbia University Press; 1942. [Google Scholar]
  16. Páeckert M, Martens J, Sun Y H. Phylogeny of longtailed tits and allies inferred from mitochondrial and nuclear markers (Ayes: Passeriformes, Aegithalidae) Mol Phylogenet Evol. 55:952–967. doi: 10.1016/j.ympev.2010.01.024. [DOI] [PubMed] [Google Scholar]
  17. Payne R B, Del Hoyo J, Elliott A, Sargatal J. Cuckoos (Cuculidae) Handbook of the birds of the world. 4:508–608. [Google Scholar]
  18. Payne R B. The cuckoos. Oxford): Oxford University Press; 2005. [Google Scholar]
  19. Peters J L. Harvard University Press; 1940. Check-list of Birds of the World. vol. 4.4 [Google Scholar]
  20. Rankin G, Stokes M. Reliability of assessment tools in rehabilitation: an illustration of appropriate statistical analyses. Clin Rehabil. 12:187–199. doi: 10.1191/026921598672178340. [DOI] [PubMed] [Google Scholar]
  21. Sibley C G, Monroe B L. Distribution and taxonomy of birds of the world. New Haven): Yale University Press; 1990. [Google Scholar]
  22. Slabbekoorn H, Smith T B. Bird song, ecology and speciation. Phil Trans R Soc Lond B. 357:493–503. doi: 10.1098/rstb.2001.1056. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Additional file 1.

Recordings used in this study and original measurement data. (download)

zoolstud-55-038-s001.xls (39.5KB, xls)

Articles from Zoological Studies are provided here courtesy of Biodiversity Research Center, Academia Sinica

RESOURCES