Molecular sexing in the formation of pairs of blue-and-yellow macaw (Ara ararauna) in reintroduction programs

Blue and yellow macaw is a species which does not show sexual dimorphism and is threatened by animal traffic. The identification of heterosexual pairs is important for reintroduction programs. The aim of this work was select parameters for sexing and use them to determine the frequency of heterosexual pairs in a population of blue-and-yellow macaws allocated in a Wild Animal Screening Center. Blood samples from 23 macaws were collected and genomic DNA extracted by Tris/SDS washes. Allele-specific molecular markers for sexing were amplified by PCR, and identified on 2% agarose gel. Three pairs of primers were tested: Pair 1 (P2/P8), Pair 2 (1237L/1272H) and Pair 3 (2550F/2718R). For the determination of animal pairs, all individuals had their social behavioral acts observed. The results showed that the low complexity DNA extraction protocol used was adequate. Pairs 2 and 3 of primers were effective for sexing and the Pair 3 was the most efficient. The study also showed that in the sample studied, the composition of males and females was similar (0.4 males n=10 and 0.6 females n=13); 70% (n=16) of the individuals formed pairs and 75% (n=12) of the pairs were heterosexual and the others male-male or female-female pairs. These results were used in the management of the animals in the reintroduction program.

However, despite the brazilian biodiversity of birds, according to the Red List of Threatened Species (2015), Brazil is considered the country with the largest number of wild bird species threatened with extinction (apud Piacentini et al., 2015). Several species of birds, because they have exuberant colors and vocalization ability, contribute to this group of animals being considered the most sought after for illegal trade (Sick, 1997).
Among the groups of birds illegally traded, the order Psittaciformes can be cited (Souza & Soares-Filho, 2005).
Psittaciformes, currently, is composed of three families: (i) Cacatuidae, family of cockatoos; and (ii) Psittacidae (which is divided into the subfamilies Loriinae and Psittacinae) and (iii) Strigopidae, composed of endemic species from New Zealand (Joseph, Toon, Schirtzinger, Wright & Schodde, 2012). With representatives of small size (such as parakeets) to large wings (such as macaws), psittacines belong to the family Psittacidae. Psittacidae also has a marked presence in Brazil, with 87 species (Piacentini et al., 2015).
The blue-and-yellow macaw are among the largest birds of the Psittacidae family, varying between 71 and 90 centimeters in length (Rodríguez-Mahecha, Suárez, Arzuza & Hernández, 2005) and are found mainly in the north of South American and also in the south of Central America (BirdLife International, 2018). A common behavior in species of the Psittacidae family, which includes the blue-and-yellow macaws, is social monogamy, in the reproductive season pairs are formed that remain throughout life (Sick, 1997).
In 1999, Bagemihl, analyzed sexual behavior in 450 species of birds and mammals, concluding that there are variations of the standard behavior established between males and females. Among the birds, these variations were observed in several species, from courtship to the formation of pairs composed of animals of the same sex. The establishment of homosexual pairs in birds has also been documented in monogamous species (Jankowiak, Tryjanowski, Hetmański & Skórka, 2018).
In birds, the apparent sexual dimorphism established by morphological characteristics is often due to feather coloration and body size (Sick, 1997). Many bird species do not exhibit marked sexual dimorphism (Sick, 1997). Among them are parrots (Allgayer & Cziulik, 2007), which include the blue-and-yellow macaw (Ara ararauna). Considering that this species does not present sexual dimorphism, the use of technical resources for sexing becomes indispensable in the process of pair identification.
In order to guarantee the reproductive success of the species in captivity or the management in the release of birds in reintroduction projects, the identification of the pairs formed by the animals should be considered.
Molecular markers, associated with the identification of the alleles present on the sex chromosomes Z and W, can be amplified by PCR using different primer pairs (Ágh, Kovács, Nemesházi & Szabó, 2018;Franco-Gutiérrez, Álvarez-Cardona & Soto-Calderón, 2017;Ong & Vellayan, 2008;Sulandari & Zein, 2012). The objective of the present study was to select parameters of molecular sexing of blue-and-yellow macaw and the use of these to determine the frequency of heterosexual pairs in the population of wild blue-and-yellow macaw allocated in one Wildlife Screening Center (CETAS) of Brazil.

Methodology
The animals in the present study were 23 adult blue-and-yellow macaw (A. ararauna), allocated in CETAS-VDC, Traditional Knowledge), registration number ACF8348. This is a descriptive quantitative research (Dalfovo, Lana & Silveira, 2008;Pereira, Shitsuka, Parreira & Shitsuka, 2018). All animals studied were identified with aluminum washers (these are cited as CETASUFBA in the Results item). Blood samples were collected from all the individuals studied, being obtained through a small cut on the nails of the macaws, using specialized pliers. The blood was collected on strips of 110 mm filter paper (JProlab®), previously cut out under sterile conditions, and allowed to dry for usage in DNA extraction, following the procedures described by Smith and Burgoyne (2004).
DNA extraction was performed using the protocol also described by Smith and Burgoyne (2004) with modifications.
We did not used 5 M guanidine thiocyanate or 95% ethanol in the washes, we increased the time in the water bath from 10 minutes to 20 minutes and we used 4μL of eluant, instead of 1μL.
The filter paper strips with dry biological material (blood) were washed using 100 mM Tris-0.1% SDS solution, with gentle and uninterrupted shaking for about 30 minutes. Then the strips with the biological material were again washed three times with Milli-Q water (ultrapure) under the same stirring conditions mentioned for the first wash. Finally, after discarding the water used in the washes, 50 μl of ultrapure water was added to the microtubes (with the strips), and these were kept in a water bath for 20 minutes. The strips were then discarded and the eluant (4μL) (containing the extracted DNA) stored. Qualitative and quantitative analyzes of DNA were performed using NanoDrop 2000® (Thermo Scientific, 44 MA, USA). 1uL of each sample is used for analysis.
Observational ad libitum methodology (Altmann, 1974) was used to identify the formation of pairs. The identification was through continuous observation (four months, four days a week, four hours a day, totaling 256 hours) of the behavior of the macaws in the enclosures where they were being kept. The enclosure was 12 m long x 6 m wide x 5.50 m high, all sides delimited by a screen, three perches and a tree. The individuals had free access to water and food (pieces of banana, papaya, apple, orange, cucumber and sunflower seeds). The identification of the pairs (heterosexual and homosexual) occurred through the observation of four social behavioral acts (Table 1) for at least three continuous weeks per pair, previously described in the literature (Favoretto, 2016). These results were used in the management of the animals in the release of bird in reintroduction program. Table 1. Social behavioral acts used to identify the heterosexual and homosexual pairs of blue-and-yellow macaw.
Behavioral acts Details (i) Beak to Beak contact birds fit one beak to the other and perform delicate movements by touching the tongues or just touching the beaks. (ii) Allopreening or Mutual allopreening the bird cleans another individual's feathers, especially in regions that are difficult to clean, such as the head, neck and cloaca region. The behavior can be mutual between the partners. (iii) Food and/or object sharing the animal moves forward, downward, backward and upward with its head, causing the food to move from the ingluvium to the beak (undigested food).
Then it moves from beak to beak and delivers the food to the other. When observing the partner manipulating an object, the bird approaches gently and with its beak tries to remove the object from the other, who can accept to share it. (iv) Staying together without sign of quarreling or annoyance.
In Table 1, complete descriptions of the determinant behaviors of peers among individuals can be seen. Such behavioral criteria, described by Favoretto (2016), were considered as indicative of the formation of pairs in the present study if observed between two individuals for at least 3 continuous weeks.

Results
The results of the analysis of extracted DNA were satisfactory, in qualitative and quantitative terms, enough to allow the amplification of the molecular markers of all individuals to be sexed. The analysis revealed a significant variation in the concentration of DNA samples obtained (DNA concentration -average 94.32 ng/μL, lowest value 3.1 ng/μL; largest value 246 ng/μL). The mean values of the sample indicators were 1.72 for organic contaminants (lower value 1.33, higher value 1.84) and 1.17 for protein contaminants (lower value 0.6, higher value 1.74).
To test the relative efficiency of the three pairs of primers, four samples were selected from the 23 samples extracted.
It was found that, even though different amounts of DNA were used and the DNA purity indicators of all samples used had been below the ideal established limits, all samples had the molecular markers amplified by PCR.
The results illustrated in Figure 1 show the visualization of amplified molecular markers. The use of molecular weight allowed to estimate the approximate sizes of the molecular markers ( Table 2).   Size difference between amplified fragments (bp) -20 170 Source: Authors.
The estimated sizes shown in Table 1 were: (i) for the Pair of primers 1: band Z -390 bp, the band W was not observed. The size differences between the molecular markers generated by both Pair 2 (approximately 20 bp) and Pair 3 (approximately 170 bp) were sufficient to allow the separation of the molecular markers on agarose gel, and consequently for the sexing of the individuals. Pair 3 showed better sexing performance. This performance is due to the Pair of primers 3 having sizes of allelespecific bands with the largest difference in size between each other, thus being better separated in agarose gel.
Three annealing temperatures were used in the thermocycling program for the Pair of primers 3, with the intention to verify which one would allow better visualization of the bands in agarose gel. At 55 °C (Figure 1), a smaller number of nonspecific bands were observed; but the allele-specific bands were visualized in a lower intensity, when compared with the allele-specific bands obtained with the temperature of 53 °C. The temperature of 53 °C ( Figure 2) amplified the allele-specific bands with higher intensity, facilitating their visualization; although nonspecific bands were also observed. At 51 °C, nonspecific bands were generated; whereas allele-specific bands were visualized with less intensity, when compared with the bands obtained at a temperature of 53 ºC. It is understood here by specific bands those with estimated sizes of 630 bp for the Z marker and 460 bp for the W allele. Thus, the present work establishes 53 °C as the best temperature to be used in the thermocycling program for amplification of the molecular markers Z and W of blue-and-yellow macaw with the Pair of primers 3.  The main modification of the DNA extraction protocol used in this work in relation to the protocol described for the use of FTA cards (Whatman ®), was the replacement of the FTA card for sterile filter paper. It does not contain substances that promote cell lysis. Vieira, Coelho and Oliveira (2009) performed sexing on 10 species of birds (including species of the family Psittacidae, but not including macaws) involving the use of three DNA extraction protocols: (i) Procedure used by Sambrook, Fritsch and Maniatis (1989), involving the use of phenol / chloroform and proteinase K enzyme; (ii) Simple and rapid alkaline extraction protocol, described by Rudbeck and Dissing (1998) for human DNA samples (blood and semen); and (iii) Modified alternative protocol from the Manual of preparation of samples using FTA cards. According to Vieira et al. (2009), the third protocol mentioned here was as efficient as the others were, but at a reduced cost. Regarding the quantity and quality parameters evaluated in the DNA extracted in the present study, it was not possible to compare with data in the literature because there are few articles mentioning it.

Best pair of primers for molecular sexing
The three pairs of primers were used to determine which would be most suitable for molecular sexing. Using Pair of primers 1, it was possible to visualize only one band (estimated size 390 bp, equivalent to the Z allele). It appeared in only one of the samples, this being the individual CETAS UFBA 013 (genotyped as male). We believe that either (i) or there was no amplification of the molecular markers from the DNAs of the other three individuals tested for amplification with the primer pair 1, (ii) or the amplified fragments were not generated in sufficient quantity for all bands to be visualized on agarose gel. A possible explanation for the difference in the performance of primer pair 1 between the results of the articles mentioned in this text and those obtained in the present work is the DNA quality of the samples used. It is worth remembering that the works compared here were used of DNAs obtained through different DNA extraction protocols.
Amplified DNA with Primer Pair 1 (P2 / P8) Griffiths et al. (1998), studying owl Strix aluco, reports that it was not possible to visualize difference between the two allele-specific bands in 3% agarose gel. Griffiths et al. (1998) Here are cited some articles that worked with allele-specific molecular markers for molecular sexing generated by Pair of primers 3 (2550F / 2718R). Fridolfsson and Ellegren (1999) working with 11 orders of birds (including the order Psittaciformes, but not including blue-and-yellow macaw) and using 3% agarose gel, observed differences between allelespecific bands of 150 bp for Gallus gallus. Thanou et al. (2013) (Phalacrocoracidae) on 4% agarose gel and eventually acrylamide 16%, observing a mean difference of 250 bp between allele-specific fragments. Vucicevic et al. (2012) performed sexing with 58 species of birds (including blue-and-yellow macaw) using a 2% agarose gel, observing a difference between 150 and 250 bp between the molecular markers of each species studied. Çakmak et al. (2016) studying 77 species of birds (including blue-andyellow macaw), using capillary electrophoresis, observed a mean difference between 265 bp allele-specific bands.
Molecular markers indicating the Z and W alleles generated by the same pair of primers with the greatest size difference between them are more efficient in distinguishing the sex genotypes after gel electrophoresis, which is especially relevant when using agarose gel. This gel, when compared to acrylamide gel, is considered less resolutive. However, the agarose gel shows significant advantages, such as easier and quicker to prepare, less costly and less toxic.
From the sexing studies in birds, the ones that were shown to be closer (by the studied species and techniques used) and that  Table 3 shows the approximate sizes of the allele-specific bands obtained in the molecular sexing of blue-and-yellow macaw and biguas (family Phalacrocoracidae) using Primers 1, 2 and 3, considering data from Jensen et al. (2003); Thanou et al. (2013) and Çakmak et al. (2016). Table 3. Estimated sizes of allele-specific molecular markers generated from three pairs of primers in the sexing of blue-andyellow macaw A. ararauna, in other works. 250 -*Estimated sizes of bands obtained for blue-and-yellow macaw, by Jensen et al. (2003). **Banding sizes obtained for bivalve family Phalacrocoracidae, by Thanou et al. (2013). ***Size of the Z band obtained for blue-and-yellow macaw, by Çakmak et al. (2016). Source: Survey done by the authors.

Allele-specific molecular marker
Observing the data in Table 3 it can be said that as for Pair of primers 1, the Z allele generated on this research was estimated (390 bp) as having the same size of the Z allele for blue-and-yellow macaw estimated by Jensen et al. (2003), Differences in size between allele-specific fragments can be explained (i) by different accuracy of fragment size estimation methods (whether in plate electrophoresis or capillary electrophoresis, and in acrylamide or agarose gel) and / or (ii) by genomic differences between species or varieties of the same species that underwent molecular sexing. The present work prioritizes the use of agarose as it is less toxic, quicker to prepare and cheaper. Such advantages are significant for studies of wild animal populations. Regarding the amplified molecular markers being allele-specific, the images of the gels and the statements of some of the articles mentioned here lead us to believe that all were specific. These results, not completely in agreement with the estimated size of the amplified molecular markers, call attention to how research in this area allows variations in results.

Best Thermocycling Program for Molecular Sexing
The molecular markers Z and W had already been amplified using the Pair of primers 3 (2550F / 2718R) by: (i) Vucicevic et al. (2012), (ii) Thanou et al. (2013) and (iii) Çakmak et al. (2016). Comparing with the three articles mentioned here, the article by Vucicevic et al. (2012) was the one that worked with more similar parameters of the present study: sexing of blue-and-yellow macaw and use of electrophoresis on agarose gel plaque. For this reason, we opted to optimize in the present work the conditions of amplification of the molecular markers of blue-and-yellow macaw using the Pair of primers 3 from the protocol of Vucicevic et al. (2012).
According to Vucicevic et al. (2012), there was amplification of only allele-specific bands at the temperature of 55ºC, described by the authors as Protocol 3 (P3). The present work showed that although using thermocycling at 55 ºC, there is a decrease in the amplification of nonspecific bands, at 53 ºC we obtain specific bands more visible in gel. Non-specific bands are easily disregarded at sexing due to sizes not coinciding with the expected sizes of the molecular markers used for sexing. Thus, we recommend thermocycling for the molecular sexing of blue-and-yellow macaw using Pair of primers 3 with Tm at 53 ºC, instead of the other possible temperatures.

Sex ratio of blue-and-yellow macaw in captivity
The prevalence of the sexes found in the sample of blue-and-yellow macaw in the CETAS-VDC was 0.6 females and 0.4 males. We believe that this proportion is representative not only of the sample used, but also of the prevalence of the sexes in macaws (all blue-and-yellow macaw) of the institution that housed them. The 23 macaws studied in the present study were randomly selected from 34 blue-and-yellow macaw allotted in the CETAS, all located in the same enclosure. Caparroz, Guedes, Bianchi and Wajntal (2001) found sexual proportion for blue-and-yellow macaw wild (0.44 females and 0.56 males) similar to that of the present article. According to the authors, these differences are not statistically significant in relation to the ratio of 1:1.

Occurrence of heterosexual and homosexual pairs in captive blue-and-yellow macaw
The blue-and-yellow macaw has monogamous reproductive behavior, forming lifelong pairs that are only reformulated in the absence of one of the partners. According to Allgayer and Cziulik (2007), captive psittacines exhibit characteristic behaviors that may be indicative of paired formation. For pairs to be formed among psittacines in captivity, there must be compatibility between the two individuals involved so that, they can live in harmony. It is possible to detect through observation that animals prone to join in pairs demonstrate mutual interest, as described by their characteristic behaviors (Allgayer & Cziulik, 2007).

Favoretto (2016) described behavioral patterns, including behaviors indicating the constitution of pairs for Lear's macaw
Anodorhynchus leari in captivity. When comparing the indicative patterns of behavior to identify pairs proposed by Favoretto (2016) with those by Allgayer and Cziulik (2007), Favoretto (2016) proposes more types of behavior patterns. These include: (i) Beak to beak contact, (ii) Allopreening or mutual allopreening, (iii) Food and/or object sharing, and (iv) Staying together without sign of quarreling or annoyance. These were the ones adopted by the present work for the identification of pairs. Bagemihl (1999) documented the composition of pairs involving individuals of the same sex in about 450 species of mammals and birds. In a review by MacFarlane, Blomberg and Vasey (2010), 83 species of wild birds showed occurrence of homosexual pairs. Dagg (1984) describes that the composition of homosexual pairs in mammals occurs most often in captivity.
Bagemihl in 1999, describes pairs made up of individuals of the same sex are found in free life and captivity, but more frequently in animals kept in captivity. Allgayer and Cziulik (2007) also describe how common, in collective enclosures, the formation of pairs between birds of the same sex. Bagemihl (1999) reports that in birds there are species in which there are individuals entering simultaneously and / or in sequence in heterosexual and homosexual pairs, which is observed especially in monogamous species. Mills (1994), reports individuals from the family Laridae (family of gulls and scissor beaks), consisting of monogamous species, composing female-female pairs. Hardy (1963) and Buchanan (1966) performed work with Eupsittula canicularis parakeets, describing the formation of female-female pairs kept in captivity (Hardy, 1963) and male-male pairs in wild population (Buchanan, 1966).
In the present study, the combination of sexing and observation of the individuals revealed that 70% of the sample of individuals in the CETAS-VDC enclosure were pairs, and 75% of the pairs were heterosexual pairs. Of the 25% of homosexual pairs observed, one was composed by males and the other by females. The above-mentioned sex ratio, not far from 1:1, suggests that heterosexual pairs were not formed due to lack of individuals of the opposite sex. The following are reports on the frequency of homosexual pairs in groups of animals or species in particular. Bagemihl (1999) described the frequency of 20% of homosexual pairs among 450 individuals (including birds and mammals), considering wild animals and kept in captivity. Kotrschal, Hemetsberger and Weiss (2006), found that in the monogamous goose-goose Anser anser 12% to 20% of pairs in the wild population were individuals of the same sex. (Young, Zaun & VanderWerf, 2008), described in a wild albatross population Phoebastria immutabilis, a frequency of 31% of homosexual pairs, with formation of only female-female pairs. When comparing the frequency of homosexual pairs found in the sample of the present study (25%) with those published in the literature, a similar frequency is observed.
We emphasize that the present study has its importance also linked to the management process necessary for the reintroduction of the species to nature. The frequency of heterosexual pairs is important to ensure the reproductive success of species both in captivity and in reintroduction programs. The reproductive success of the species is directly related to the frequency of homosexual pairs, since the female-female pairs imply the non-fertilization of the eggs and the male-male pairs involve the non-egg laying. In other words, homosexual pairs can bring low fertility rates in the communities that have them . Therefore, the frequency of heterosexual and homosexual pairs has a direct impact on the reintroduction programs, including programs promoted by CETAS; which aim to ensure the reproductive success of individuals in the environment in which they are reintroduced. The research conducted by Sanz and Rodriguez-Ferraro (2006), for example, reaffirm the necessity to prioritize the reproductive success of the species, referring to Amazona barbadensis both in captivity and in the environment in which they will be reintroduced.
No information was found in the scientific literature on the composition (nor the frequency) of same sex pairs in macaws and blue-and-yellow macaw (wild or in captive). Considering the above, even considering sample of individuals in captivity and relatively limited sample size (n = 23), we believe that this work brings significant contributions to the understanding of the biology and behavior of macaws, especially of the species A. ararauna.

Conclusion
Therefore, the present study demonstrated that in relation to the selection of parameters for molecular sexing of blueand-yellow macaw (i) the low complexity DNA extraction protocol used was adequate; (ii) Pairs of primers 2 and 3 were effective for sexing; (iii) and the Pair of primer 3 was the most efficient. The study also showed that in the sample of wild blue-and-yellow macaw studied in CETAS-VDC, (iv) the sample composition of males and females was similar (0.4 males n = 10 and 0.6 females n = 13), (v) 70% (n = 16) of the individuals were distributed in pairs, (vi) being 75% (n = 12) of the heterosexual pairs, and (vii) having male-male and female-female pairs.
Our data reinforce the importance to use molecular biology along with conservation monitoring. Through assertive and accessible technics, this work elucidated the formation of pairs among blue-and-yellow macaw improving the reintroduction strategy of these animals. We hope that these results encourage and provide new insights about reintroduction of species that do not present sexual dimorphism. These results can also help the development of conservation measures of A. ararauna and contribute to future research about general bird reintroduction in nature by monitoring pairs formation.