Molecular sexing and stable isotope analyses reveal incomplete sexual dimorphism and potential breeding range of Siberian Rubythroats Luscinia calliope captured in Taiwan

Abstract

The Siberian Rubythroat Luscinia calliope breeds widely across Siberia and several other north Asian countries, migrating to overwinter to the south, including Taiwan. Its presence in Taiwan presented a unique opportunity to investigate questions which could not be reliably answered before the development of modern molecular techniques. We used molecular sexing techniques to determine whether there is overlap between the sexes in the measurements of colour and morphometric traits. We also used stable hydrogen isotope analysis to determine the potential breeding areas of these individuals. We found consistent morphometric and colour differences between males and females, but with so much variation that no single measurement could be used to sex individuals and, as already suspected, neither size nor colour of the red throat-patch or the bordering sub-malar stripes are reliable field marks for sexing Siberian Rubythroats. However, a combination of wing length and red throat-patch area when entered into a logistic regression correctly identified all males and females. This result could not be repeated when we used a discriminant function with cross-validation, which is currently the standard procedure for sex discrimination. The stable hydrogen isotope analysis showed that individuals captured in Taiwan potentially originate from large parts of its known breeding range, but mostly from southern areas. This suggests a 'leap-frog' migration by this species. Furthermore, larger individuals originated from higher latitudes, in accordance with Bergmann's rule. Modern molecular techniques thus shed interesting new insights into the morphology and ecology of Siberian Rubythroats.

Full text

INTRODUCTION
The Siberian Rubythroat Luscinia calliope breeds from the central
Ural Mountains across Siberia to the coast of Far East Russia,
northern Japan, northern Korea, central China and parts of
Mongolia, and overwinters in South and South-East Asia, the
south-east seaboard of China, Taiwan and the Philippines (Glutz
von Blotzheim 1988, Pagenkopf 2003, Collar 2005, Brazil 2009).
In Taiwan, Siberian Rubythroats are passage migrants or stay to
overwinter (Wang et al. 1991); ringing data collected by the Chinese
Wild Bird Federation (CWBF) between 2000–2007 showed that
the species arrives in Taiwan in October and returns north by early
May.
The presence of Siberian Rubythroats in Taiwan presented us
with a unique opportunity to investigate some questions which
could not be reliably answered before the arrival of modern
molecular techniques: (1) are Siberian Rubythroats consistently
dimorphic, or is there overlap in the measurements of colour and
morphometric traits between the sexes? (2) where are the breeding
areas of Siberian Rubythroats which overwinter in Taiwan?
Due to the rapid advance of reliable molecular techniques
(Griffiths et al. 1998, Fridolfsson & Ellegren 1999), sexing of birds
has moved from using colour, morphometric or behavioural
differences to more reliable DNA techniques (e.g. Jodice et al. 2000,
Redman et al. 2002). Before these techniques were available, the
main colour difference between Siberian Rubythroat sexes was
described as the metallic ruby-red chin and throat of the male
compared with the white chin and throat of the female (Kennedy
et al. 2000, Severinghaus et al. 2010), although most field guides
and handbooks noted that some females had a yellow-tinged buff
(Collar 2005), pinkish or even reddish throat (Glutz von Blotzheim
1988, Barthel 1996, Grimmett et al. 1998, Lee et al. 2000,
MacKinnon & Phillipps 2000, Bergmann 2001, Robson 2008,
Brazil 2009). However, it is unclear from these sources how reliable
the sexing of individuals was; most are field guides, while others
(Glutz von Blotzheim 1988, Barthel 1996, Collar 2005) cite
original research but their sources all pre-date the use of DNA
sexing techniques.
Likewise, Pagenkopf (2003) documented that 28% of breeding
females in central Siberia had obvious red colouration on their
throats, and then cited Svensson (1992) which claimed that 80%
of females showed such colouration. While Svensson (1992)
determined the sex using body size, brood-patch and the shape of
the cloacal region, Pagenkopf (2003) did not describe how he
verified the sexes, but claimed that a better diagnostic colour trait
than the throat colour is the dark lines bordering the throat-patch
(hereafter referred to as ‘sub-malar stripes’), supposedly always black
in males and brown in females. As a result of these conflicting
identification guidelines, there are many records of unsexed Siberian
Rubythroats in the literature (e.g. Pagenkopf 2003) and databases
(e.g. the CWBF Taiwan bird banding database). Females are, on
average, smaller than males (Glutz von Blotzheim 1988, Kennedy
et al. 2000, Pagenkopf 2003). Pagenkopf (2003) established that
females, on average, had less body mass and shorter wings, tails,
tarsi and toes (with all differences statistically significant at the p <
0.001 level, except for the back toe); this assertion was supported
by his review of 16 previous publications (Pagenkopf 2003).
To our knowledge, no previous study had used DNA sexing
techniques when we adopted this method to clarify the extent and
variation of the supposed colour and morphometric differences
between male and female Siberian Rubythroats. Because DNA
sexing techniques require invasive blood sampling and extensive
laboratory work, we also investigated whether we could develop a
statistical method based on morphometric and colour traits to
reliably determine the sex of Siberian Rubythroats.
Due to the rapid advance of stable-hydrogen isotope techniques,
we can now determine the potential breeding or wintering areas of
migratory species, at least to the accuracy of a regional scale (e.g.
Hobson 2005, Reichlin et al. 2010). Since birds consume water
directly or indirectly via food, the stable-hydrogen isotope ratios
of their feathers (δDf) should reflect the δD of the precipitation
(δDp) of the geographic region where an individual grows its
feathers (Hobson & Wassenaar 1997). So far, only a few studies
(Rocque et al. 2006, 2009, Chang et al. 2008, Pérez et al. 2010)
have used this technique in East Asia, and none for the Siberian
Rubythroat. In the case of some species, linkage between breeding
Molecular sexing and stable isotope analyses reveal
incomplete sexual dimorphism and potential breeding
range of Siberian Rubythroats Luscinia calliope
captured in Taiwan
GUO-JING WENG, HUI-SHAN LIN, YUAN-HSUN SUN & BRUNO A. WALTHER
The Siberian Rubythroat Luscinia calliope breeds widely across Siberia and several other north Asian countries, migrating to overwinter to
the south, including Taiwan. Its presence in Taiwan presented a unique opportunity to investigate questions which could not be reliably
answered before the development of modern molecular techniques. We used molecular sexing techniques to determine whether there is
overlap between the sexes in the measurements of colour and morphometric traits. We also used stable hydrogen isotope analysis to
determine the potential breeding areas of these individuals. We found consistent morphometric and colour differences between males and
females, but with so much variation that no single measurement could be used to sex individuals and, as already suspected, neither size nor
colour of the red throat-patch or the bordering sub-malar stripes are reliable field marks for sexing Siberian Rubythroats. However, a
combination of wing length and red throat-patch area when entered into a logistic regression correctly identified all males and females.
This result could not be repeated when we used a discriminant function with cross-validation, which is currently the standard procedure for
sex discrimination. The stable hydrogen isotope analysis showed that individuals captured in Taiwan potentially originate from large parts
of its known breeding range, but mostly from southern areas. This suggests a ‘leap-frog’ migration by this species. Furthermore, larger
individuals originated from higher latitudes, in accordance with Bergmann’s rule. Modern molecular techniques thus shed interesting new
insights into the morphology and ecology of Siberian Rubythroats.
FORKTAIL 30 (2014): 96–103

and wintering areas was established through ringing recoveries,
satellite tracking or morphological studies (e.g. using different
races). However, Siberian Rubythroats have very low ringing
recovery rates, are too small to carry transmitters, and show little,
if any, racial variation (Glutz von Blotzheim 1988, Barthel 1996,
Pagenkopf 2003, Collar 2005). Therefore, we used the stable-
hydrogen isotope technique to determine the potential breeding
areas of Siberian Rubythroats captured in Taiwan.
METHODS
Capture and handling of individuals
Siberian Rubythroats were sampled during winter 2007/2008.
Between 6 November and 14 December 2007, birds were examined
in a cage-bird trade shop in Neipu township, Pingtung county,
south-west Taiwan; these had been captured by using playback to
lure them into mist-nets set along the banks of the Donggang River
and its tributaries (22.633°N 120.617°E). We arranged to be
immediately informed when birds were captured and, consequently,
individuals were sampled, measured, banded and released within
3 days or less of their arrival at the shop. Between 21 January and
27 February 2008, birds were trapped using the same technique
around the headwaters of the Nioujiaowan stream (22.633–
22.650°N 120.417–120.583°E) upstream of the Donggang River.
Essentially, both sets of birds were captured in the same area using
the same method but by two different groups of people.
We assumed that the measured colour and morphometric traits
and δDf values did not change during captivity. However, we did
exclude body weight from our analyses. Blood and feather samples
along with colour and morphometric measurements were taken
from each individual when possible. However, a few individuals
developed signs of stress (e.g. open bills and fast breathing) during
handling and were released immediately, hence the unequal sample
sizes in some of our analyses.
Molecular sexing
A few drops of blood were collected from one brachial vein of each
individual using half-inch needles and capillaries and then stored
in 99% ethanol until isolation of genomic DNA. We isolated total
genomic DNA using traditional chloroform/isoamyl alcohol (24:1)
extraction and then identified the sex of each individual with the
2550F (5'-GTTACTGATTCGTCTACGAGA-3') and 2718R
(5'-ATTGAAATGATCCAGTGCTTG-3') primer set
(Fridolfsson & Ellegren 1999). We performed PCRs on iCycler
thermal cyclers (Bio-Rad) in 10-
μ
L reaction volumes containing
about 0.5 ng of genomic DNA, 0.1 mM of each dNTP, 1 × PCR
buffer (BioScience), 0.1 U of Ta q DNA polymerase (BioScience),
0.5 mM of MgCl2, and 0.2
μ
M of each primer. The PCR conditions
were 95°C for 180 s, followed by 35 cycles of 95°C for 30 s, 46°C
for 40 s, and 72°C for 100 s. The final extension was at 72°C for
300 s. PCR products were separated in 1.2% agarose gels and
visualised by ethidium bromide staining. To minimise risk of error,
the entire procedure was repeated independently by two different
laboratory workers.
Age determination
Each individual was classified as subadult (first-winter) or adult.
Subadults have incompletely ossified skulls ( Jenni & Winkler 1994)
and/or light-brown tips on the greater coverts 1–7, and tertials
which c ontrast with the dark brow n of the remainder of the feathers
(Glutz von Blotzheim 1988, Barthel 1996, Collar 2005). However,
subadults may lose the light-brown tips through abrasion (Glutz
von Blotzheim 1988) and/or may have completely ossified skulls.
Therefore, individuals with either light-brown tips or incompletely
ossified skulls were classified as subadults; all others were classified
as adults. To ensure consistency, all age classification, colour and
morphometric measurements were conducted by H-SL.
Colour and morphometric traits
All morphometric measurements (Table 1) were taken according
to instructions detailed in banding manuals of the Taiwan Bird
Banding Center (1989) and were measured to the nearest 0.1 mm,
except for the area of the red throat-patch which was determined
by counting the red-filled 2.5×2.5 mm grid squares in photographs
of each individual (Plate 1). When grid squares were only partially
red, each was classified visually using ten categories (red coverage
1–10%, 11–20% etc). Any ‘orange’ areas were scored as red because
the ‘orange’ is the result of a different viewing angle or a change in
ambient light. Although several authors suggested that the sub-
malar stripes are black in males and more grey-brown in females
Table 1. Colour and morpholometric traits of Siberian Rubythroats.
Two-tailed Mann-Whitney U tests were used to compare differences between males and females.
Males Females
Morphometric traits nMin Max Med Mean s.d. nMin Max Med Mean s.d. U-value p-value
Red throat-patch area (mm2) 39 202.5 413.1 276.9 283.5 45.1 17 0 272.5 70.6 116.8 105.8 193 <0.0001
Head length (mm) 39 36.2 39.8 38.2 38.2 0.8 16 35.9 38.9 37.2 37.3 0.9 272.5 0.001
Bill length (mm) 37 11.5 14.3 13.3 13.2 0.7 16 12.2 14.0 13.0 12.9 0.5 331 0.050
Body length (mm) 39 141.0 159.0 151.0 149.9 4.3 16 135.0 152.0 146.5 145.5 5.0 294 0.004
Tarsus length (mm) 39 29.3 32.8 30.8 30.9 0.9 17 28.6 30.9 29.5 29.7 0.7 255 <0.0001
Wing length (mm) 39 74.5 83.5 79.0 78.8 2 17 72.0 77.0 75.0 75.1 1.3 198 <0.0001
Flattened wing length (mm) 39 75.5 86.5 81.0 80.5 2.2 17 75.0 79.5 77.0 77.0 1.3 212.5 <0.0001
Tail length (mm) 39 58.0 68.0 62.5 62.4 2.7 16 55.0 63.5 60.0 59.7 2.3 276.5 0.002
Plate 1. Variation of red throat-patches and sub-malar stripes of
Siberian Rubythroats. (A) A typical male with dark red throat-patch and
black sub-malar stripes. Clear plastic with a grid size of 2.5×2.5 mm
was placed over the throat to estimate the area covered by red colour.
(B) An atypical female showing significant red throat-patch and strong
sub-malar stripes. (C) A typical female with a faint reddish throat-patch
and faint light-brown sub-malar stripes. A sub-malar stripe in B and
the throat-patch in C were manually framed (thin yellow lines) with
the software ImageJ to measure their grey values.
Forktail 30 (2014) Molecular sexing and stable isotope analyses of Siberian Rubythroats Luscinia calliope in Taiwan 97

98 GUO-JING WENG et al. Forktail 30 (2014)
(see Introduction), we noticed that the darkness of the sub-malar
stripes correlated with the darkness of the throat-patches, regardless
of sex. In order to quantify the darkness of each throat-patch and
sub-malar stripe, a front-view photo of each individual was taken.
Grey values of the throat-patch and sub-malar stripe were measured
under the red, green and blue (RGB) mode using the software
ImageJ (Schneider et al. 2012). The grey value measures the light
intensity at each pixel with black as the weakest (grey value = 0)
and white at the strongest (grey value = 255) intensity. The areas
of each throat-patch and each sub-malar stripe to be measured were
manually framed using the Freehand Sele ction tool in Imag eJ (Plate
1). For each throat-patch, the entire area was used, including any
whitish areas. For the sub-malar stripes, only one was randomly
chosen and measured, as we assumed that the light intensity of the
two stripes was essentially equal.
Stable-hydrogen isotope analyses
We combined maps (Glutz von Blotzheim 1988, Grimmett et al.
1998, Collar 2005) of the breeding and wintering areas of Siberian
Rubythroat to create a shape-file of maximum breeding and
wintering areas using ArcGIS 9.2 software. To interpret the
geographic locations of the breeding area of Siberian Rubythroats
correctly, we needed to ensure that the collected feather samples
were actually grown in the breeding region and not later during
migration or wintering. We used three approaches to evaluate the
reliability of the feather samples.
First, we considered the published moult chronology of
Siberian Rubythroats to sample the appropriate feathers from which
to analyse δDf. After the breeding season but before migration, adult
Siberian Rubythroats go through a complete annual moult between
August and September, while subadults moult the feathers of the
head, body, lesser coverts, medium coverts and parts of the greater
coverts 8–10 between late August and late September (Glutz von
Blotzheim 1988, Svensson 1992, Cramp 1998). Therefore, for
adults and subadults, the δDf values of the rectrices should reflect
the δDp of the breeding region. We collected the innermost rectrix
on the right side (R1) from each individual. In one case, R1 had
been lost for an unknown reason. Because the replacement feather
growing at R1 could therefore not be used, we collected the first
rectrix on the left side of R1. We refer to these feathers as ‘breeding
feathers’.
Second, it is well established that subadults grow all of their
feathers on the breeding grounds prior to migration. Therefore,
similar mean δDf values of the breeding feathers for subadults and
adults would further indicate that the breeding feathers of adults
are also grown in the breeding area and not later during migration.
On the same lines of reasoning, the δDf values of the breeding
feathers should exhibit smaller variation if they were grown in the
breeding area and larger variation if they were grown along the
migration route. We thus compared the mean and variation of δDf
values between subadults and adults.
Third, we collected feathers from nine individuals showing signs
of recent feather growth. We refer to these feathers as ‘wintering
feathers’. If any of our breeding feathers were actually grown during
migration, we would expect to observe δDf values for at least some
individuals which would lie somewhere between the mean δDf
value of the breeding and wintering feathers. If the arguments made
above all favour the hypothesis that all breeding feathers were
indeed grown on the breeding grounds, we could then be confident
that we could use the δDf values of our breeding feathers to interpret
the geographic range where Siberian Rubythroats breed.
To determine the distribution of δDp values within the
maximum breeding and overwintering areas, we downloaded the
global mean δDp map of the growing season from http://
www.waterisotopes.org (Bowen et al. 2005); the resolution of its
grid squares was 20×20 geographic minutes. This growing-season
δDp map (Figure 3) shows clear and roughly latitudinal δDp bands
within the maximum breeding range, without any trend from coast
to inland, which suggests that heavier (and thus less negative) δDp
values are roughly associated with lower latitudes. Consequently,
the δDf values of the breeding feathers should also reflect this
latitudinal patterns if the conditions above hold.
The mean δDp value for Pingtung county from November to
the following March was -20‰ during 2005 and 2007 (C.-H. Wang
in litt. 2013) while the mean δDf value of our wintering feathers
was -37‰, which is 17‰ more depleted than the δDp value (see
Results). This difference of -17‰ was used as the discrimination
factor to transform the growing-season δDp map to the map of
expected δDf values (Figure 4).
Fixed discrimination factors have been widely used for assigning
migratory species to their breeding area, e.g. Northern Pintails Anas
acuta breeding in Alaska (Yerkes et al. 2008) where water sources
are probably as variable as in the breeding area of Siberian
Rubythroats. Similarly, Pérez et al. (2010) assigned waterbirds
moulting in north-central Mongolia to their breeding region using
a fixed discrimination factor of -28‰ calculated for North
American waterfowl by Clark et al. (2006). We also used a
discrimination factor but, unlike Pérez et al. (2010), we derived it
from our own data, specifically -17‰ as explained above.
We thus subtracted 17‰ from each grid square of our growing-
season δDp map (Figure 3) to derive a map of the p otential breed ing
area of Siberian Rubythroats (Figure 4) within the maximum
breeding range. We then colour-labelled those grid squares whose
value matched the mean δDf value of our breeding feathers as well
as those grid squares which fell (1) within one standard error, (2)
within the 95% confidence interval of the mean δDf value and (3)
within the total range of the δDf of the breeding feathers, to rank
each grid square for its likelihood as the location where these
feathers were actually grown.
After collection, all feathers were soaked in a 2:1
chloroform:methanol solution to remove surface oil and dirt; then
each sample was air-dried and then weighed to 0.5 mg before
analysis (Wassenaar & Hobson 2000, Norris et al. 2006). The
samples were analysed at the Stable Isotope Facility, University of
California, Davis. There, feather samples and four keratin standards
with known values of non-exchangeable hydrogen (BWB, CFS,
CHS and CH1) were stored in a vacuum desiccator to equilibrate
the exchangeable hydrogen in feather samples and standards. The
Heckatech HT Oxygen Analyzer, interfaced to a continuous-flow
Isotope Ratio Mass Spectrometer (IRMS) known as PDZ Europa
20-20 (Sercon Ltd., Cheshire, UK), was used for the analysis during
which the keratin standards and the standard VSMOW (Vienna
Standard Mean Ocean Water), which has no exchangeable
hydrogen with -120‰ (NBS22), were used to adjust for the stable-
hydrogen isotope ratios of the feather samples.
Statistical analyses
Since some data relating to colour, morphometric and feather
measurements were not normally distributed, the Mann-Whitney
U test was used to test for differences between two groups, while
the Pearson’s correlation coefficient (r) was used to evaluate
correlation between two variables. We also used a non-parametric
one-sample sign-test to test for a general trend in the direction of
correlations between latitude and measures of body size. We used
SPSS 12.0 for these analyses.
To choose the optimal model for sex determination of Siberian
Rubythroats, we used the logistic regression procedure with the
stepwise effect-selection method in SAS 9.0, whereby the binary
dependent variable sex was regressed against nine independent
variables, namely the seven morphometric traits (Table 1), the red
throat-patch area (Table 1) and the grey value (Figure 1).
Independent variables were added to the model when p < 0.05 and

were excluded from the model when p > 0.35. Once some of these
independent variables had been selected by this logistic regression
model, these same variables were then used in a discriminant analysis
for sex determination. We also conducted a cross-validation to
evaluate the performance of the discriminant function. The
individual to be validated was not included when we calculated
the respective discriminant f unction, and this discriminant function
was then used to determine the sex of this omitted individual.
RESULTS
Sex determination using DNA analyses
A total of 56 individuals (8 from mist-nets, 48 from the cage-bird
trade shop) were captured. Genetic sexing identified 39 males and
17 females (Table 2), with the results from the two independent
laboratory workers agreeing completely.
Morphometric and colour traits
No significant differences in morphometric or colour traits (Table
1) were found between the birds captured by the authors (n = 8)
and by the cage-bird shop owners (n = 48) except for tail length—
cage-bird shop birds had marginally longer tails (U = 136.0, p =
0.04). We therefore lumped all individuals together. All medians
and means of our eight traits were greater for male than for female;
moreover, all differences were statistically significant at the p < 0.05
level except for bill length (Table 1). However, there was always
some overl ap: the max imum female mea surement was alway s greater
than the minimum male measurement but always smaller than the
maximum male measurement. The minimum female measurement
was always smaller than the minimum male measurement, except
for bill length (Table 1).
Classifying the throat-patch colouration into three types, we
observed that 43 of the 56 individuals (76.8%) had red throat-
patches (Table 2). All males fell into this category, but also one
subadult and three adult females. The nine individuals with red-
and-white throats (16.1%) and the four with white throats (7.1%)
were all females (Table 2). Thus, if individuals with red patches
had been identified exclusively as males and individuals with red-
and-white patches or white patches exclusively as females, 7.1% of
all individuals and 23.5% of all females would have been incorrectly
sexed. Twenty-eight subadult males (100%) but only one subadult
female (10%) had red throat-patches (Table 2), indicating that, at
least for males, the red throat-patch appeared in their first winter
and was not a good indicator of age at this time of year, in
accordance with Barthel (1996).
In both sexes the average area of the red throat-patch was larger
in adults than in subadults, but these differences were not
statistically significant in either sex (male: U = 271.0, p = 0.11;
female: U = 65.5, p = 0.81), which is due to the large overlap in
values (Figure 2). The grey value of the throat-patch was positively
correlated with the grey value of the sub-malar stripe (n = 56, r =
0.73, p < 0.0001; Figure 1). This relationship was slightly stronger
for females (n = 17, r = 0.72, p = 0.002) than for males (n = 39, r =
0.44, p = 0.005). Figure 1 further illustrates that, in general, males
have significantly darker throat-patches and sub-malar stripes than
females but again with some overlap (U = 668 and 742, p = 0.001
and < 0.0001, respectively). Variation of grey values was larger in
females than in males for throat-patch (s.d. = 45.2 and 19.9
respectively) and for sub-malar stripe (s.d. = 32.2 and 22.4
respectively) (Figure 1). In both sexes, there was no statistically
significant difference for the grey values of the throat-patch or the
sub-malar stripe between subadults and adults (Mann-Whitney
U tests, all p > 0.08; Figure 1).
Table 2. Variation of the colour of the throat-patch among sex and
age categories in Siberian Rubythroats. The red and white categories
meant that > 95% of the throat-patch was either red or white, and all
other individuals fell into the red-and-white category.
Colour of throat-patch
Sex Age Red Red-and-white White Total
Male Adult 11 0 0 11
Subadult 28 0 0 28
Female Adult 3 3 1 7
Subadult 1 6 3 10
Total 43 9 4 56
Figure 1. Relationship between the grey values of the red throat-patch
and the grey values of the sub-malar stripe of Siberian Rubythroats
(filled circle = adult male, open circle = subadult male, closed triangle
= adult female, open triangle = subadult female). Possible grey values
range from 0 (black) to 255 (white), whereby the corresponding
gradation bars show that lower grey values appear darker in vision.
The lower-left and upper-right crosses show the means and standard
deviations for males and females, respectively.
Figure 2. Relationship between the red throat-patch area and the wing
length of Siberian Rubythroats (same symbols as in Figure 1). The
values of two females overlap at zero red throat-patch area and 75
mm wing length.
Forktail 30 (2014) Molecular sexing and stable isotope analyses of Siberian Rubythroats Luscinia calliope in Taiwan 99

100 GUO-JING WENG et al. Forktail 30 (2014)
The logistic regression identified two independent variables
which correlated with the binary dependent variable sex, namely
red throat-patch area and wing length (Table 3). This function
yielded a 100% accuracy of sex determination because negative
values resulting from the application of the logistic regression
function always identified males correctly, and positive values
always females correctly. Acc ordingly, plotting red throat-patch area
versus wing length shows clearly separated clusters of males and
females, although it also shows the considerable overlap between
the sexes for both variables when they are considered by themselves
(Figure 2).
Using red throat-patch area and wing length as independent
variables, a quadratic discriminant function was established which
misclassified two males and one female in cross-validation, which
amounts to an error rate of 5.1% in identifying males and 5.9% in
identifying females. The cross-validation failed to identif y sexes as
perfectly as the logistic regression model because the two
misidentified males had relatively short wing lengths (75 mm) while
the misidentified female had a relatively long wing length (77 mm).
The cross-validation thus showed that male and female Siberian
Rubythroats could not always be reliably identified within a
discriminant function.
Isotope analysis
The δDf values of the breeding feathers did not differ significantly
between the sex and age categories with an overall mean ± s.d. of
-99.9‰ ± 12.0‰ (Table 4; sex comparison: U = 540.0, p = 0.64;
Table 4. Sample size (n), mean ± standard deviation, minimum and
maximum δDf values of the breeding feathers collected from Siberian
Rubythroats of different ages and sexes.
δδ
δδ
δDf (‰)
nMean ± s.d. Minimum Maximum
Sex Male 39 -100.9 ± 11.6 -126 -72
Female 17 -97.6 ± 13.0 -118 -72
Age Adult 18 -99.2 ± 10.5 -119 -80
Subadult 38 -100.3 ± 12.8 -126 -72
Total 56 -99.9 ± 12.0 -126 -72
Table 3. An example of one of the possible logistic regression models
to distinguish the sexes of Siberian Rubythroats. Because a gap exists
between males and females in Figure 2, a maximum likelihood estimate
does not exist and therefore the logistic equation is not unique; rather,
many different possibilities exist, of which the one below is one
example. Negative values always identify males and positive values
always identify females correctly (number of observations = 47).
Parameter F Estimate SE Wald chi-square p-value
Intercept 1 705.60 597.70 1.394 0.24
Red throat-patch area 1 -0.36 0.31 1.425 0.23
Wing length 1 -8.08 6.90 1.372 0.24
Figure 3. Distribution of the δDp values across the published maximum breeding region (outlined in blue) of the Siberian Rubythroat.
age comparison: U = 527.5, p = 0.44). Indeed, the mean δDf values
of the breeding feathers from adults and subadults were almost
identical (Table 4), suggesting that all individuals originated from
similar regions (see Methods). Furthermore, if any of the breeding
feathers were grown outside the breeding range, we would see a
larger s.d. of δDf for adults than for subadults (see Methods), but
adults possessed a smaller s.d. than subadults (Table 4). Finally, our
wintering feathers (mean ± s.d.: -37.0‰ ± 7.7‰, range: -45‰ to
-23‰) differed substantially from the breeding feathers in terms
of mean, s.d., and range of δDf values (Table 4). If any of the
breeding feathers had been grown during migration, we should have
found δDf values somewhere between -99.9 and -37.0‰, as
explained in the Methods. These facts thus support the hypothesis
that all breeding feathers were indeed grown in the breeding area
and can be used to pinpoint the bree ding region using the growing-
season δDp map (Figure 3) and our discrimination factor.
The derived map of the potential breeding areas of Siberian
Rubythroats overwintering in Taiwan shows that, except for the
most westerly part of the published breeding range, the potential
breeding area identified by the most likely grids is located mostly
in the southern part of the maximum breeding range (Figure 4).
As estimates of overall body size, we chose five of our seven
morphometric traits (Table 1). We randomly excluded one of the
two measurements of wing length and excluded b ill length because
it is not well correlated with overall body size. Correlating these
five measurements with the mean δDf values of the breeding

feathers, there is a general trend for larger individuals to have lower
mean δDf values (Table 5), which are roughly associated with higher
latitudes (Figure 4).
DISCUSSION
Through the use of relatively novel molecular techniques, we were
able for the first time to reliably test how male and female Siberian
Rubythroats differ in colour and morphometric traits as well as to
determine the potential breeding grounds of individuals captured
in Taiwan.
Our results demonstrate that neither the size nor the colour of
the red throat-patch nor the bordering sub-malar stripes are reliable
field marks for sexing Siberian Rubythroats, particularly females.
Whilst males overall have larger and darker throat-patches and
darker sub-malar stripes, there is overlap between males and females
and therefore scope for misidentification. The same is true for all
morphometric traits which cannot be used in isolation to reliably
sex individuals. We emphasise that these results are novel because
this is the first study based on reliable sexing techniques. Therefore,
previous studies (Glutz von Blotzheim 1988, Svensson 1992,
Barthel 1996, Cramp 1998, Pagenkopf 2003) which described
colour and morphometric differences between the sexes should now
be viewed cautiously, as some individuals may have been
misidentified if sexing was not based on additional sexing methods,
e.g. song, brood-patch or sexual organs.
While a logistic regression without cross-validation was able to
reliably sex individuals, this result could not be repeated with a
discriminant function analysis, which is the standard procedure for
sex discrimination. Further studies should test whether these results
would change when sample sizes are increased or when different
populations are sampled.
Previous studies have claimed that fledglings within their first
year of life have no or almost no red in their throat-patches,
regardless of sex (Glutz von Blotzheim 1988, Bergmann 2001),
although Barthel (1996) differed in claiming that subadult males
already have red throats during the autumn of their first year while
maintaining their light-brown feather tips. Barthel’s claim is
supported by our findings that both subadult and adult males have
red throat-patches, while subadult and adult females may have
white, red-and-white or red throat-patches (Table 2).
Bergmann (2001) claimed that the red throat colouration is
dependent on nutrition and possibly exposure to sunlig ht, as captive
birds lose most/all of the throat’s red c olouration after their moult.
This effect is well known from other species (Hudson 1994). If age
is not an important determining factor for the area and darkness of
the red throat-patch, as our results suggest, then perhaps the red
throat-patch is a signal which indicates the capability of its bearer
to obtain the required carotenoid-containing foods (Olson &
Owens 1998, Senar & Escobar 2002, Saks et al. 2003, Griggio et
al. 2007).
Despite the conclusion that the colour and morphometric traits
discussed above cannot by themselves (but possibly in combination)
reliably sex individuals in every case, our study lends support to
previous studies which revealed that, on average, males are larger
Table 5. Correlations between five body size measurements and mean
δDf values of breeding feathers taken from Siberian Rubythroats
captured in Taiwan (r = Pearson’s correlation coefficient; p = p-value;
n = sample size). Eighteen of 20 correlations were negative (one sample
sign-test, p = 0.0004), meaning that there was an overall significant
trend that larger individuals originated from higher latitudes (which
are associated with lower δDf values), even if only two of the 20
correlations themselves had an associated p-value < 0.05.
Male Female Adult Subadult
Head length r -0.17 0.08 -0.01 -0.17
p0.29 0.77 0.96 0.32
n39 16 17 38
Body length r -0.09 -0.51 0.02 -0.22
p0.60 0.04 0.95 0.20
n39 16 18 37
Tarsus length r -0.18 -0.004 -0.32 -0.12
p0.28 0.99 0.20 0.47
n39 17 18 38
Flattened wing length r -0.38 -0.05 -0.40 - 0.30
p0.02 0.85 0.10 0.07
n39 17 18 38
Tail length r -0.13 -0.44 -0.29 -0.25
p0.42 0.09 0.24 0.13
n39 16 18 37
Figure 4. Map of the potential
breeding area of Siberian
Rubythroats overwintering in
Taiwan. The published maximum
overwintering and breeding
ranges of the Siberian Rubythroat
are outlined in light blue and blue,
respectively. Only those grid
squares within the maximum
breeding range which were equal
to the mean δDf value of the
breeding feathers (dark green),
within one standard error (green),
within the 95% confidence
interval (light green) around the
mean, or within the total range of
δDf values of the breeding
feathers (pale green), have been
coloured.
Forktail 30 (2014) Molecular sexing and stable isotope analyses of Siberian Rubythroats Luscinia calliope in Taiwan 101

102 GUO-JING WENG et al. Forktail 30 (2014)
and have larger and darker red throat-patches and darker sub-malar
stripes. Therefore, besides the obvious colour dimorphism of the
throat area, there is indeed also a slight size dimorphism in the
species. Owens & Hartley (1998) found that differences in size
dimorphism were associated with variation in social mating system
and sex difference in parental care, while differences in plumage-
colour dimorphism were associated with variation in the frequency
of extra-pair paternity. Thus social mating system, parental care and
extra-pair paternity could all play a role in the evolution of the
Siberian Rubythroat’s body size and colours. The closely related
Bluethroat Luscinia svecica has already been used to investigate the
relation between colouration and male and female mate choic e, mate
guarding and sperm competition (Amundsen et al. 1997, Johnsen
et al. 1998, 2003). In the future, Asian ornithologists should perhaps
consider using the Siberian Rubythroat to study these topics.
Since the mean and variation of our δDf values of the breeding
and wintering feathers met the conditions set out in our Methods,
we are confident that the breeding feathers were all grown within
the breeding area. Furthermore, the s.d. values of our breeding
feathers (Table 4) were comparable with those reported from
feathers grown at similar latitudes by other bird species examined
in Europe (Hobson et al. 2004); 25 species, s.d. = 1.0–10.6‰) and
North America (Hobson & Wassenaar 1997); 6 species, s.d. = 4.0–
11.0‰), indicating that Siberian Rubythroats overwintering in
Taiwan orig inated from a relatively restricted latitudinal range. Based
on their studies in North America, Farmer et al. (2008) stated that
‘two samples that differ by less than 31‰ cannot be confidently
said to originate from different latitudes’. The δDf values of our
breeding feathers ranged from -126‰ to -72‰, thus spanning
approximately 54‰, which means that our samples should span
about 2° of latitude. The relatively small s.d. and range of our δDf
values thus suggest that the actual breeding range of Siberian
Rubythroats overwintering in Taiwan could be even more restricted
than Figure 4 suggests.
Our map of the potential breeding area of Siberian Rubythroats
(Figure 4) suggests that birds overwintering in Taiwan originate
only from the southern part of the published breeding range, except
for the most westerly regions. Given that Taiwan is located at the
most northerly part of the documented overwintering range, this
result suggests possible leap-frog migration (Kelly et al. 2002,
Paxton et al. 2007, Reichlin et al. 2010), whereby the most northerly
breeding populations migrate to the most southerly overwintering
regions and thereby literally leap over those populations which
migrate the least latitudinal distance from breeding to
overwintering regions. Therefore, it would be interesting to repeat
this analysis for other overwintering populations within east Asia
to reliably establish the linkages between various breeding and
overwintering areas.
Given that there is a rough correlation between δDp and
latitude (Figure 3), our results (Table 5) also suggest that larger
individuals breed at higher latitudes. Such a correlation is in
agreement with Bergmann’s rule (Bergmann 1847, Futuyma 2009).
However, as discussed above, our Siberian Rubythroats probably
originated from within only 2° of latitude. This finding should
therefore also be corroborated by further studies of other
overwintering populations within east Asia.
Modern molecular techniques, in combination with older
techniques such as colour and morphometric measurements, thus
shed interesting new insights into the morphology and ecology of
Siberian Rubythroats.
ACKNOWLEDGEMENTS
We thank Chung-Ho Wang for providing the hydrogen isotopic composition
of the precipitation in Pingtung county, Yu- Cheng Hsu for assisting in DNA
sex determination, Andrew Gosler, Klaus Michalek, Roger Riddington, Susan
Samuel and Norbert Schδffer for providing references, and Tsai-Yu Wu for
help with translation and with information on migratory birds in Taiwan.
BAW was financially supported by a grant from Taipei Medical University.
REFERENCES
Amundsen, T., Forsgren, E. & Hansen, L. T. T. (1997) On the function of female
ornaments: male Bluethroats prefer colourful females. Proc. R. Soc. Lond.
B 264: 1579–1586.
Barthel, P. H. (1996) Die Bestimmung des Rubinkehlchens Luscinia calliope
und sein Auftreten in Europa. Limicola 10: 180–189.
Bergmann, C. (1847) Über die Verhältnisse der Wärmeökonomie der Thiere
zu ihrer Grösse. Göttinger Studien 3: 595–708.
Bergmann, H.-H. (2001) Das Rubinkehlchen. Falke 48: 183–185.
Bowen, G. J., Wassenaar, L. I. & Hobson, K. A. (2005) Global application of
stable hydrogen and oxygen isotopes to wildlife forensics. Oecologia
143: 337–348.
Brazil, M. (2009) Birds of east Asia. Princeton: Princeton University Press.
Chang, Y.-M., Hatch, K. A., Ding, T.-S., Eggett, D. L., Yuan, H.-W. & Roeder, B.
L. (2008) Using stable isotopes to unravel and predict the origins of
great cormorants (Phalacrocorax carbo sinensis) overwintering at
Kinmen. Rapid Communications in Mass Spectrometry 22: 1235–1244.
Clark, R. G., Hobson, K. A. & Wassenaar, L. I. (2006) Geographic variation in
the isotopic (δ D, δ 13C, δ 15N, δ 34S) composition of feathers and claws
from lesser scaup and northern pintail: implications for studies of
migratory connectivity. Can. J. Zool. 84: 1395–1401.
Collar, N. J. (2005) Family Turdidae (thrushes and chats). Pp.514–807 in J.
del Hoyo, A. Elliott & D. A. Christie, eds. Handbook of the birds of the
world, 10. Barcelona: Lynx Edicions.
Cramp, S. (1998) The complete birds of the Western Palearctic on CD-ROM.
Oxford: Oxford University Press.
Farmer, A., Cade, B. & Torres-Dowdall, J. (2008) Fundamental limits to the
accuracy of deuterium isotopes for identifying the spatial origin of
migratory animals. Oecologia 158: 183–192.
Fridolfsson, A.-K. & Ellegren, H. (1999) A simple and universal method for
molecular sexing of non-ratite birds. J. Avian Biol. 30: 116–121.
Futuyma, D. J. (2009) Evolution. Second edition. Sunderland Mass: Sinauer
Associates.
Glutz von Blotzheim, U. N. (1988) Handbuch der Vögel Mitteleuropas: Band
11/I. Passeriformes (2. Teil): Turdidae. Schmätzer und Verwandte:
Erithacinae. Wiesbaden: Aula-Verlag.
Griffiths, R., Double, M. C., Orr, K. & Dawson, R. J. G. (1998) A DNA test to sex
most birds. Molec. Ecol. 7: 1071–1075.
Griggio, M., Serra, L., Licheri, D., Monti, A. & Pilastro, A. (2007) Armaments
and ornaments in the rock sparrow: a possible dual utility of a
carotenoid-based feather signal. Behav. Ecol. Sociobiol. 61: 423–433.
Grimmett, R., Inskipp, C. & Inskipp, T. (1998) Birds of the Indian subcontinent.
London: Christopher Helm.
Hobson, K. A. (2005) Using stable isotopes to trace long-distance dispersal
in birds and other taxa. Div. Distrib. 11: 157–164.
Hobson, K. A. & Wassenaar, L. I. (1997) Linking breeding and wintering
grounds of neotropical migrant songbirds using stable hydrogen
isotopic analysis of feathers. Oecologia 109: 142–148.
Hobson, K. A., Bowen, G. J., Wassenaar, L. I., Ferrand, Y. & Lormee, H. (2004)
Using stable hydrogen and oxygen isotope measurements of feathers
to infer geographical origins of migrating European birds. Oecologia
141: 477–488.
Hudson, J. (1994) Showiness, carotenoids, and captivity—a comment on
Hill (1992). Auk 111: 218–221.
Jenni, L. & Winkler, R. (1994) Moult and ageing of European passerines.
London: Academic Press.
Jodice, P. G. R., Lanctot, R. B., Gill, V. A., Roby, D. D. & Hatch, S. A. (2000)
Sexing adult black-legged kittiwakes by DNA, behavior, and
morphology. Waterbirds 23: 405–415.

Johnsen, A., Andersson, S., Ornborg, J. & Lifjeld, J. T. (1998) Ultraviolet plumage
ornamentation affects social mate choice and sperm competition in
Bluethroats (Aves: Luscinia s. svecica): a field experiment. Proc. R. Soc. Lond.
B 265: 1313–1318.
Johnsen, A., Lifjeld, J. T. & Krokene, C. (2003) Age-related variation in mate-
guarding intensity in the Bluethroat (Luscinia s. svecica). Ethology 109:
147–158.
Kelly, J., Atudorei, V., Sharp, Z. & Finch, D. (2002) Insights into Wilson’s Warbler
migration from analyses of hydrogen stable-isotope ratios. Oecologia
130: 216–221.
Kennedy, R. S., Gonzales, P. C., Dickinson, E. C., Miranda, H. C. & Fisher, T. H.
(2000) A guide to the birds of the Philippines. Oxford: Oxford University
Press.
Lee W. S., Koo T. H. & Park J. Y. (2000) A field guide to the birds of Korea. Seoul:
LG Evergreen Foundation.
MacKinnon, J. & Phillipps, K. (2000) A field guide to the birds of China. Oxford:
Oxford University Press.
Norris, D. R., Marra, P. P., Bowen, G. J., Ratcliffe, L. M., Royle, J. A. & Kyser, T. K.
(2006) Migratory connectivity of a widely distributed songbird, the
American redstart (Setophaga ruticilla). Orn. Monogr. 61: 14–28.
Olson, V. A. & Owens, I. P. F. (1998) Costly sexual signals: are carotenoids rare,
risky or required? Trends Ecol. Evol. 13: 510–514.
Owens, I. P. F. & Hartley, I. R. (1998) Sexual dimorphism in birds: why are there
so many different forms of dimorphism? Proc. R. Soc. Lond. B 265: 397–
407.
Pagenkopf, K. (2003) Vergleichende Studien an sibirischen Nachtigallen der
Gattung Luscinia am mittleren Jenissei – Zentralsibirien. Münster: PhD,
Westfälische Wilhelms-Universität Münster.
Paxton, K. L., Van Riper, C., Theimer, T. C. & Paxton, E. H. (2007) Spatial and
temporal migration patterns of Wilson’s warbler (Wilsonia pusilla) in the
southwest as revealed by stable isotopes. Auk 124: 162–175.
Pérez, G. E., Hobson, K. A., Garde, E. J. & Gilbert, M. (2010) Deuterium (δD) in
feathers of Mongolian waterbirds uncovers migratory movements.
Waterbirds 33: 438–443.
Redman, K. K., Lewis, S., Griffiths, R., Wanless, S. & Hamer, K. C. (2002) Sexing
northern gannets from DNA, morphology and behavior. Waterbirds 25:
230–234.
Reichlin, T. S., Hobson, K. A., Wassenaar, L. I., Schaub, M., Tolkmitt, D., Becker,
D., Jenni, L. & Arlettaz, R. (2010) Migratory connectivity in a declining
bird species: using feather isotopes to inform demographic modelling.
Div. Distrib. 16: 643–654.
Robson, C. (2008) A guide to the birds of South-East Asia. London: New Holland.
Rocque, D. A., Ben-David, M., Barry, R. P. & Winker, K. (2006) Assigning birds to
wintering and breeding grounds using stable isotopes: lessons from two
feather generations among three intercontinental migrants. J. Orn. 147:
395–404.
Rocque, D. A., Ben-David, M., Barry, R. P. & Winker, K. (2009) Wheatear molt
and assignment tests: ongoing lessons in using stable isotopes to infer
origins. J. Orn. 150: 931–934.
Saks, L., Ots, I. & Hõrak, P. (2003) Carotenoid-based plumage coloration of
male greenfinches reflects health and immunocompetence. Oecologia
134: 301–307.
Schneider, C. A., Rasband, W. S. & Eliceiri, K. W. (2012) NIH Image to ImageJ:
25 years of image analysis. Nat. Methods 9: 671–675.
Senar, J. C. & Escobar, D. (2002) Carotenoid-derived plumage coloration in
the siskin Carduelis spinus is related to foraging ability. Avian Sci. 2:
19–24.
Severinghaus, L. L., Ding T.-S., Fang W.-H., Lin W.-H., Tsai M.-C. & Yen C.-W.
(2010) The avifauna of Taiwan. Taipei: Forestry Bureau of Council of
Agriculture of Executive Yuan. (In Chinese.)
Svensson, L. (1992) Identification guide to European passerines. Fourth
edition. Thetford: British Trust for Ornithology.
Taiwan Bird Banding Center (1989) A guide to banding birds in Taiwan. Taipei:
Chinese Wild Bird Federation of Taiwan. (In Chinese.)
Wang C., Wu S., Kuang K., Ying C., Chei Z. & Siao C. (1991) A field guide to
wild birds of Taiwan. Taipei: Yasher Press. (In Chinese.)
Wassenaar, L. I. & Hobson, K. A. (2000) Improved method for determining
the stable-hydrogen isotopic composition (δD) of complex organic
materials of environmental interest. Environ. Sci. Technol. 34: 2354–2360.
Yerkes, T., Hobson, K. A., Wassenaar, L. I., Macleod, R. & Coluccy, J. M. (2008)
Stable isotopes (δD, δ13C, δ15N) reveal associations among geographic
location and condition of Alaskan northern pintails. J. Wildl. Mgmt. 72:
715–725.
Guo-Jing WENG, Institute of Wildlife Conservation, National
Pingtung University of Science and Technology, Pingtung 912,
Taiwan. Email: gjw@mail.npust.edu.tw
Hui-Shan LIN, Institute of Wildlife Conservation, National
Pingtung University of Science and Technology, Pingtung 912,
Taiwan. Email: blackkite588@yahoo.com.tw
Yuan-Hsun SUN, Institute of Wildlife Conservation, National
Pingtung University of Science and Technology, Pingtung 912,
Taiwan. Email: ysun@mail.npust.edu.tw
Bruno A. WALTHER, Master Program in Global Health and
Development, College of Public Health and Nutrition, Taipei
Medical University, 250 Wuxing St., Taipei 110, Taiwan. Email:
bawalther2009@gmail.com (corresponding author)
Forktail 30 (2014) Molecular sexing and stable isotope analyses of Siberian Rubythroats Luscinia calliope in Taiwan 103