Sex determination in Turdus amaurochalinus ( Passeriformes : Muscicapidae ) : morphometrical analysis supported by CHD gene

Sex determination is important for conservation and population studies, particularly for reproduction programs of threatened species and behavioural ecology. Turdus amaurochalinus, Creamy-bellied Thrush, only exhibits sexual dimorphism during the breeding season, when males are considered to show intense yellow bills, and females and immature males show dark brown bills. The objectives of this study were: 1) to determine the sex of individuals using genetic techniques, and 2) to test the hypothesis that sex dimorphism can be detected by morphometry. This study was carried out at Parque Nacional da Restinga de Jurubatiba, a preserved area located on the North coast of Rio de Janeiro State. The birds were captured using ornithological nets, singly marked with metal rings, weighed, measured and had blood samples collected before being released. The sex of 42 T. amaurochalinus individuals was determined using the CHD gene marker. A total of 20 males and 22 females were identified from June to August, with peak capture frequency in June. Turdus amaurochalinus females and males differed significantly in morphometrical measures. The most important traits to distinguish males from females were wing length (Student t-test=4.34, df=40, p=0.0001) and weight (Student t-test=2.08, df=40, p=0.044): females were heavier and had significantly shorter wing length than males. Females and males were correctly classified in 86% and 75% of cases, respectively, using Discriminant Analysis. The molecular analysis was the most secure method for sex determination in the studied species. Rev. Biol. Trop. 59 (2): 789794. Epub 2011 June 01.

Many bird species cannot be sexed by any morphological trait, not differing in morphological external characteristics, while others are sexually dimorphic only in the adult phase (Ellegren & Sheldon 1997, Kahn et al. 1998).In the case of Turdus amaurochalinus, it is difficult to differentiate young males from adult females.
Sex is one of the most important variables to distinguish individuals in a population.According to Fisher's theory about sexual proportion, if equal conditions were available for male and female production, the sex ratio at the time parental investment ends should stabilize at 1:1 (Fisher 1930).The individuals of many populations diverge from the expected sex ratio of 50% for each sex, not being able to assume a 1:1 constant ratio (Krebs 1994).However, these differences between species and populations could have importance, if correlated to some particularity of their life cycle (Bull & Charnov 1988, Lens et al. 1998).
Techniques for sex determination in birds include laparoscopy, karyotyping (Basrur et al. 1998), analysis of the faecal steroid, DNA fingerprinting and molecular techniques based on the chromo-helicase-DNA binding (CHD) gene (Ellegren 1996, Griffiths et al. 1998).The CHD gene has been used successfully in many bird species (Griffiths et al. 1998, Miyaki et al. 1998, Ito et al. 2003, Sacchi et al. 2004, Lee et al. 2007, 2010), since the gene is preserved in most bird species (Griffiths & Tiwari 1995).
In birds, females are the heterogametic sex (ZW) and males are homogametic (ZZ) (Ellegren 1996).The W chromosome is special for the female, and the determination of the sex is made based on the absence or presence of the marker W-linked.Using polymerase chain reaction (PCR) techniques, it is possible to easily identify the females, since they show two bands on agarose gel.The CHD gene was successfully used for sex determination in the White-necked Thrush Turdus albicollis (Vieillot, 1818), not apparently sexually dimorphic in the adult phase.For this species, males and females differed in morphometric characteristics such as wing length and body mass, which were significantly bigger in males (Ritter et al. 2003).For this reason, we would expect to find distinct morphological characteristics between sexes in the con-generic species, Creamy-bellied Thrush T. amaurochalinus Cabanis 1850, found in the forests of Argentina, Bolivia, Brazil, Chile, Paraguay, Peru and Uruguay.
The present study aims to determine the sex proportion of T. amaurochalinus using CHD gene marker, and to evaluate if morphological characteristics that may help to identify the sex of an individual in the field.

Study sites:
The work was carried out in an area of Restinga, which is part of the Atlantic forest biome, in Restinga de Jurubatiba National Park, located at the Eastern coast of Rio de Janeiro State,.The region is dominated by a large lake of fresh water (Lagoa Feia) and by a sandy quaternary plain, which stretches towards the continent, advancing from the sea to the interior approximately 2km, where the seasonally flooded forest begins to appear.There is accented seasonality, with maximum rainfall in the summer (approximately 190mm) and minimum in the winter (around 40mm).The mean annual temperature was 22.6°C (Henriques et al. 1986).
Data collection at field: Birds were captured twice a month from 2003 to 2004 in the Restinga de Jurubatiba National Park (Alves et al. 2004).Each month 10 nets were set up in open Clusia scrub and 10 nets in Forest Formation (seasonally flooded forest), with two consecutive days in each area for each excursion (20 nets total per sample period).Nets were exposed for 7 hours of sampling each day, with half of this period in the morning (06:30-10:30h) and half in the afternoon (14:30-17:30h).
Captured birds were placed individually in clean cotton bags, and were marked separately with metal and coloured rings.Shortly after capture, we recorded the following measurements: total length, wing length, length of the tail, tarsus length, exposed culmen, nostril-tip, opening-bill base, bill height in the nostril and in the base, bill width in the nostril and in the base, and length of the head to the bill tip (Sick 1997).The morphometrical data were always taken by the same researcher (one of the authors, M.A.S. Alves) to avoid bias.
We collected approximately 50-150 l of blood from the tarsal vein using a disposable needle 13x4.5mm(26G1/2) and 50 l capillary tubes with heparin.Blood was immediately transferred to 1.5ml plastic tubes with absolute ethanol; all samples were stored at room temperature during field work and at 4°C in the laboratory.
Laboratory analysis: Sex determination was carried out at the Bird Ecology Laboratory, Ecology Department of the Universidade do Estado do Rio de Janeiro (UERJ) using the CHD gene technique developed by Griffiths et al. (1998), and modified by Miyaki et al. (1998).DNA was extracted using the phenolchloroform extraction/alcohol precipitation method described in detail by Bruford et al. (1992) or by cellular lysis as described by Khatib & Gruenbaum (1996).
Initially, the CHD gene sex determination technique was tested for several bird species, including T. amaurochalinus.Some individuals of T. albicollis were used as positive controls for the method (Ritter et al. 2003) and Ramphocelus bresilius (Linnaeus, 1766) as positive controls for sex differentiation.This last species presents obvious sexual dimorphism of plumage and iris colour: males, including the young males, present red iris colour, and females, chestnut iris colour (Nogueira & Alves 2008).The result proved that the protocol developed for this technique is secure, since R. bresilius males presented one band, while females presented two bands.
The statistical analyses of data followed Zar (1984) and were carried out in SYSTAT software (version 10.2).Only data collected from T. amaurochalinus adults, selected by plumage, tarsus and bill commissure characters, were used.Discriminant Analysis includes only the variables with observations for all forty individuals: body mass, total length, wing length, length of the tail, tarsus length, exposed culmen and length of the head to the bill tip.

Sexual reason and morphometric analyses:
We categorized 42 (20 males and 22 females) T. amaurochalinus: males and females differed significantly (Test-t, p<0.05;Table 1) on two of 11 analyzed variables.Males presented longer wing length than females, while females appeared heavier than males (Fig. 1).Discriminant Analysis showed that 86% of the females and 75% of the males were categorized correctly (Wilk's lambda=0.549and p-tail=0.0028).The variables that contributed more to distinguish the sexes were wing length and exposed culmen (Table 2).The males have longer wing length than females, while females presented a bigger value for exposed culmen.
We effectively amplified a selected region of the CHD-W gene and its homologous copy CHD-Z from T. amaurochalinus with P2/P8 primers (Griffiths et al. 1998).PCR products were 380/340 base pairs (bp) in length (CHD-W/CHD-Z, respectively).

DISCUSSION
It is reported that during breeding period, T. amaurochalinus males show intense yellow upper bill, and females and immature males show black coloured upper bill (Sick 1997).However, this correlation was not demonstrated in this study, since females were recorded with partially yellow bills.
Just wing length and body mass variables were statistically significant to separate sexes when Student's t-test was applied (Table 1 and Fig. 1).Regarding body mass, this result is already anticipated, since capture was carried out during species' reproduction time, when females are heavier due to eggs production.The fact that males presented larger wing length can be due to their major displacement (i.e.flies longer distances) in living area regarding females.When Discriminant Analysis was applied to T. amaurochalinus morphometrical data, 86% of the females and 75% of the males were classified correctly.Wing length was the most significant variable in sex distinction of this species, with exposed culmen the second most significant variable (Table 2).
Student's t-test and Discriminating Analysis ratified wing length as a very significant variable to sexually separate individuals.Lens et al. (1998) and Ritter et al. (2003) also concluded that wing length is a relevant factor to separate adult males from females in Thorichthys helleri and Turdus albicollis, respectively.
Turdus amaurochalinus males presented wing length larger than females, while body mass was larger in females than in males.These characters can be used as indication of sex status in field.However, molecular analysis remains the most accurate method to determine sex for this species.

Fig. 1 .
Fig. 1.Student's t-test analysis for weight (A) and wing length (B) variables in females (F) and males (M) of Turdus amaurochalinus captured.Sex was determined using CHD gene molecular technique.

TABLE 1
Turdus amaurochalinus females and males body mass and morphometrical variables taken in the Restinga de Jurubatiba National Park, Rio de Janeiro, BrazilThe sex was determined using CHD gene molecular technique.Variables were analysed through Student's t-test (t): n=sample number, x=average, sd=standard deviation, df=degrees of freedom and p=probability.