Morphological consequences of hybridization in two interbreeding taxa: Kurdish Wheatear (Oenanthe xanthoprymna) and Persian Wheatear (O.chrysopygia) in western Iran

Department of Environment sciences, Faculty of Environment and Energy, Science and Research Branch, Islamic Azad University, P.O. Box: 14155/4933, 14515/775, Tehran, Iran. Department of Fisheries and Environmental Sciences, College of Agricultural and Natural Resources, University of Tehran, P.O. Box 4111 Karaj, Iran. Department of Biology, Faculty of Science, Ferdowsi University of Mashhad, Mashhad, Iran. Laboratoire de Zoologie Mammifères et Oiseaux, Muséum National d'Histoire Naturelle, 55 rue Buffon, F-75005 Paris, France. Service de Systématique moléculaire, CNRS FR 1541, Muséum National d'Histoire Naturelle, 43 rue Cuvier, F-75005 Paris, France. Laboratoire Ecologie et Biogéographie des Vertébrés (EPHE), Centre d'Ecologie Fonctionnelle et Evolutive, UMR 5175, 1919 route de Mende, 34293 Montpellier cedex 5, France.

Persian Wheatear is always known by its grayish under parts and red tail in both male and female coloration. In Kurdish Wheatear the male has brown-black throat and under the wings, red tail and white basal in two-thirds of rectrices, while the female is like Persian Wheatear. Persian Wheatear breeds in the inner Zagros Mountains in southwestern Iran and in the northwest, north, northeast, south and southeast (Vaurie, 1949;Cornwallis, 1975;Panov, 2005). Breeding range of the Kurdish Wheatear extends from the extreme South-eastern parts of Turkey further south-east, into the Zagros Mountains. In west and north-west of Iran, where the breeding ranges of these two taxa overlap (Figure 1), interbreeding takes place and birds with intermediate color variants (named as Oenanthe cummingi in some references) can be found (Vaurie, 1949). The characters found in this intermediate population (O. cummingi), are present in few first year birds of Kurdish Wheatear (Roselaar, 1995). Moreover these variants and Kurdish Wheatear are reported to breed side by side (Harms, 1925), probably a further proof that they belong to Kurdish Wheatear. But McCarthy (2006)   birds are sometimes lumped.
In this paper, we report on apparent hybrid Wheatears; also we tried to answer the following questions: (1) What are the morphological and plumage coloration similarities/dissimilarities of supposed hybrid with O.xanthoprymna and O.chrysopygia also with few close Oenanthe species? (2) What are the relationships between morphology, foraging method, flight method and migration in our taxa?

Sampling and field works
Field works carried out in May 2006 and 2007 in south, south west, west, northwest, center and north east of Iran and we looked for all suitable habitats for our taxa inside and outside of contact zone based on their distribution range as reported by Panov (2005). We took fifteen O.chrysopygia and ten supposed hybrid during 2 sampling years in contact zone in west (Kermanshah Province, Amrolah Region). Furthermore we took O.chrysopygia outside of contact zone in east (Northern Khorasan, 3 specimens), west (Kurdistan, 3 specimens) and center (Isfahan, 10 specimens).We could not collect O.xanthoprymna due to delays in receiving hunting permit. It seems this species had migrated to Turkey, when we arrived at the region. Because of low security, we were not able to visit this species in Iran-Turkey and Iran-Iraq borders. Instead, we used morphological measurements of eight specimens of O.xanthoprymna deposited in Tring natural history museum (UK). There are no specimens of supposed hybrid in museum collections (based on comprehensive searchs in Iran and museums in other countries). We added measurements of two close species to O.xanthoprymna and O.chrysopygia Outlaw et al., 2010); O.lugens (36 specimens) and O.finschii (20 specimens) to compare also O.alboniger (20 specimens) as outgroup from Kaboli et al. (2007a). The final data set contained 125 individuals supposed to represent 6 taxa of in Oenanthe genus

Morphometrical analyses
We took 14 external morphometrical measurements (Appendix 1) on 125 adult specimens with digital calipers to the nearest 0.2 mm following Kaboli et al (2007a, b). Measurements were made by only one person (The second author) to avoid observer bias. The final data set for the PCA and MANOVA contained 14 variables that were assigned to 3 functional groups: (i) flight apparatus (wing and tail; 6 variables), (ii) feeding apparatus (3 variables), (iii) foot-leg complex (5 variables). We calculated five ratios (secondary variables) from these primary variables (Appendix 1).

Plumage coloration analyses
We divided complete bird body (except tail) to 17 chromatic mosaics. Then, categorized chromatic characteristics of each mosaic based on the ranges of visible colors in different species of Oenanthe and allocated a color code to each mosaic. The color codes transformed to a numerical code for each mosaic to use in multiple correspondence analysis (Appendix 2). All measurements were made by only one person (The second author) to avoid observer bias. We used ADE-4 package (Thioulouse et al., 1997) for multivariate analysis and SPSS 13.0 (2007) for statistical tests.

PCA and MANOVA
We log-transformed all biometrical values (Sokal and Rohlf, 1979) to avoid problems associated with applying multivariate methods to matrices containing ratios (Atchley et al., 1976). We performed PCA on the 125× × × ×19 morphometrical matrices ("PCA19") in order to reveal patterns of correlation among variables. Also we conducted a MANOVA to test significance of differences between different groups, considering all morphometrical characters analyzed.

Morphometrical and plumage coloration distances between species
Size and shape variables were averaged for each taxon and matrices of mahalanobis distance were calculated from these mean values. Then, morphometric tree was prepared by calculating the dissimilarity among populations by average distance coefficient and by computing an UPGMA analysis. The dendrogram was rooted on two sister taxa for O.xanthoprymna and O.chrysopygia Outlaw et al., 2010); O.lugens and O.finschii to compare also O.alboniger as outgroup from Kaboli et al. (2007a).
MCA was performed on qualitative variables. We used Hill & Smith Analysis, which is a special case allowing analyzing together, a normalized PCA and MCA was used. We prepared final dendrogram by calculating the similarities/dissimilarities among populations through the average distance coefficient based on Hill & Smith analysis results.

Species in morphospace of size and shape variables
According to PCA results, three first principal components  components extracted 57% of the variation of morphological traits. PC1 which extracted 34% of the variation, was a good measure of size (correlation with the long primary feathers varies up to 0. 9) By plotting data in a morphospace, the 125 individuals of different taxa can be divided into different groups differentiated mainly by size and shape related characters. According to PC1-PC2 plane and correlation circle (Figure 2), O.xanthoprymna with long and pointed wings, long and strong bill, long tail, week feet, long tarsus and relatively long tarsus length/wing length, is clearly discriminated from other taxa by rapid and direct flight also numerous take-offs. Persian Wheatear (O.chrysopygia) and supposed hybrid had short and rounded wings, short tail and bill, short tarsus and relatively strong feet that allow species to increase their running speed and field of view (Grant, 1966). Then, we suggest the resemblance between morphometrical characters of O.chrysopygia and supposed hybrid and also their same foraging method and residence in same habitats (rocky slopes).

Dendrogram based on morphometrical and plumage coloration distances
Dendrogram based on morphometrical distances ( This was for overcoming plumage coloration characters in this tree; because it was exactly the same with the dendrogram based on only plumage coloration patterns (not shown). Therefore, we confirm the incongruence of morphometrical characters and plumage coloration patterns in our results. But as stated by Panov (2005), color patterns in wheatears are not sufficiently conservative, and should be used with great caution in looking for species relationships. Also based on , in Oenanthe, certain color characters (e.g., a black throat or a white cap) can appear, disappear and re-appear independently in different lineages (see also Price and Pavelka, 1996;Cibois et al., 2004;Olsson et al., 2005). Therefore, we based more on morphometrical variables than plumage coloration patterns.

Conclusion
In conclusion and based on morphometrical and plumage