Chromosomal similarity between the Scaly-headed parrot ( Pionus maximiliani ) , the Short-tailed parrot ( Graydidascalus brachyurus ) and the Yellow-faced parrot ( Salvatoria xanthops ) ( Psittaciformes : Aves ) : A cytotaxonomic analysis

Behavior, morphology, allozyme studies and DNA hybridization and sequencing data all suggest the independent evolution of the Old and New World parrots and support tribe status for the American species, although the phylogenetic relationships within this tribe are still poorly understood. A previous study has shown that the Yellow-faced parrot (Amazona xanthops Spix, 1824) exhibits large karyotypic differences compared to the other Amazona species and suggested that this species should be renamed Salvatoria xanthops, although the relationships between S. xanthops and the other New World parrots remain unclear. In the present work, we describe the karyotype of the Scaly-headed parrot (Pionus maximiliani Kuhl, 1820) and the karyotype and C-banding pattern of the Short-tailed parrot (Graydidascalus brachyurus Kuhl, 1820) and compare them to the karyotype and C-banding pattern of S. xanthops, as well as to the karyotypes of other New World parrots. The chromosomal similarity between these three species and the karyotypic differences between them and other New World parrots suggest that G. brachyurus and S. xanthops are sister species and are most closely related to members of the genus Pionus.


Introduction
New World parrots not only share many unique characters (e.g.chicks with imperforated ear canals and their one-legged copulatory stance) but are also wide biochemically distant from other parrots (Mainardi, 1962;Gysels, 1964), leading Smith (1975) to suggest that the New World parrots appear to form a monophyletic radiation.Allozyme (Randi and Bertagnolio, 1990) and DNA hybridization (Sibley and Alquist, 1990) studies also support the separation of Old and New World parrots, and Collar (1997) has suggested that the approximately 150 species of New world parrots should be accorded the status of a tribe (Arini).The separation of Old and New World parrots is also supported by DNA sequencing studies (Miyaki et al., 1998), although the exact phylogenetic relationships within this group are still poorly understood.
Chromosomal analyses have produced data that could help to shed light on the evolutionary and taxonomic relationships within the New World parrots.Generally, where structural differences in karyotypes exist between various species of a group it is possible to form hypotheses as to their phylogenetic relationships (Van Dongen and De Boer, 1984).The American parrots is one of the most cytogenetically studied groups of New World birds (Lucca, 1974(Lucca, , 1984(Lucca, , 1985;;De Boer and Belterman, 1980;Van Dongen and De Boer, 1984;Schmutz and Prus, 1987;Valentine, 1987;Aquino andFerrari, 1985, 1990;Duarte and Giannoni, 1990;Lucca et al., 1991;Archangelo et al., 1995;Duarte and Caparroz, 1995;Rocha et al., 1995;Goldschmidt et al., 1997;Francisco et al., 2001;Lunardi et al., 2003), the species so far analyzed exhibiting a heterogeneous karyotypic morphology among some genera in contrast with several other avian groups in which very uniform karyotypes are found.Duarte and Caparroz (1995) showed that the Yellow-faced parrot (Amazona xanthops Spix, 1824), a species endemic to eastern and central Brazil (Forshaw, 1989), exhibited large karyotypic differences not only with the other species of the genus Amazona but also all the other New World parrots that have so far been described, while the morphological analysis conducted by Sick (1984) led him to state that the `inclusion of (A. xanthops) in the Amazona genus may not be correct.Duarte and Caparroz (1995) used these differences to suggest that A. xanthops Spix, 1824 should be excluded from the genus Amazona and renamed Salvatoria xanthops (as first used by Ribeiro, 1920) to form the monotypic genus Salvatoria which is cytogenetically most closely related to the genus Pionus.Despite these findings, a clear understanding of the relationships between S. xanthops and the other New World parrots remains unclear.
In this paper we describe the karyotype of the Scalyheaded parrot (Pionus maximiliani Kuhl, 1820), which occurs from northern Argentina to the northern Brazilian states of Ceará, Piauí and Goiás (Forshaw, 1989), and the karyotype and C-banding pattern of the Short-tailed parrot (Graydidascalus brachyurus Kuhl, 1820) which inhabits the forests of Colombia, Ecuador, Peru and the Brazilian Amazon basin (Forshaw, 1989).We also discuss the phylogenetic relationships of P. Maximiliani and G. brachyurus in relation to the previously reported karyotype of S. xanthops (Duarte and Caparroz, 1995) and other New World parrots.

Material and Methods
We investigated four male and three female Graydidascalus brachyurus Kuhl, 1820, one male and two female Pionus maximiliani Kuhl, 1820 and two female S. xanthops captive parrots maintained in Zoos (see acknowledgments for details).
Metaphase chromosomes were obtained using the short-term feather-pulp culture technique as described by Duarte and Caparroz (1995).Briefly, tissue extracted from the bulbs of young feathers was cultured for six hours at 39 °C in a complete culture medium containing 8.0 mL of Mc Coy's medium, 1.0 mL of inactivated horse serum and 0.6 mL of phytohemaglutinin to which 0.0016% of colchicine was added for the last two hours.After culture, the cells were exposed to hypotonic KCl solution (0.075 M) for 15 min at 39 °C and then fixed in 3:1 methanol/acetic acid and the cell suspension dropped onto clean slides and stained with 3.5% Giemsa-phosphate (pH 6.8).
The C-banding patterns were obtained for the G. brachyurus and S. xanthops specimens (the chromosomes of the P. maximiliani specimens were not C-banded) using a modification of the barium hydroxide (Ba(OH) 2 ) method (Sumner, 1972) in which slides containing cells with metaphase chromosomes were placed in 0.01 M aqueous HCl for 20 min at room temperature and then into 5% aqueous barium hydroxide for 7 to 15 min at 37 °C followed by washing with 0.01 M aqueous HCl for 30 s to remove ex-cess barium hydroxide.After washing, the slides were placed into 2 x SSC (salina sodium-citrate solution) for one hour at 57 °C and then stained with 3.5% Giemsaphosphate (pH 6.8).
Chromosome biometry was used to determine the arm-ratio (AR) of the chromosomes which were then classification according to centromeric position using the nomenclature of Levan et al. (1964).

Results
The metaphases obtained were of good quality and the karyotypes of all the parrots sampled could be established.The diploid number was 72 for P. maximiliani and 64 for G. brachyurus, although several metaphases had a lower than expected chromosome number due to chromosomes loss and in this case the highest chromosome counts were used to establish the diploid number.
In P. Maximiliani the first seven chromosome pairs were macrochromosomes and the remaining pairs microchromosomes (Figure 1), with the first five autosomes being submetacentric (AR = 2.21, 2.39, 1,87, 1,75 and 1,72) and the sixth pair telocentric.The microchromosomes of P. maximiliani were telocentric up to the point to which their morphology could be identified, while the Z-chromosome was submetacentric (AR = 1.87) and the largest in the karyotype, and the W-chromosome was submetacentric (AR = 1.69) and equivalent in size to the autosomic pair 4.
The karyotype of G. brachyurus comprised 18 macro and 46 microchromosomes (Figure 2), with the first five autosomes being submetacentric (AR = 2.47, 2.71, 2.07, 2.17 and 1.74) and the sixth pair telocentric while pairs 7 (AR = 1.25) and 8 (AR = 1.14) were metacentric.The G. brachyurus Z-chromosome was submetacentric (AR = 2.63) and the largest in the karyotype, while the W-chromosome was submetacentric (AR = 2.26) and equivalent in size to the autosomic pair 4. Except for the presence of two small pairs (7 and 8) of metacentric macrochromosomes the chromosome morphology of G. brachyurus was very similar to that of P. maximiliani, and both species showing a very sharp boundary between macro and microchromosomes.
The C-banding patterns of G. brachyurus and S. xanthops revealed centromeric bands in almost all chromosomes (Figure 3).However, the C-bands were more pronounced in S. xanthops than in G. brachyurus.In both species, the W-chromosome presented a conspicuous pattern with the short-arm and the pericentromeric region of the long arm being entirely heterochromatic.

Discussion
Within the order Psittaciformes, 69 species belonging to 36 genera from the families Cacatuidae (9 species) and Psittacidae (60 species) have been studied cytologically.Forty-one species of Psittacidae are from the New World Caparroz and Duarte and cytological studies have revealed at least three main karyotypic patterns in the New world parrots.
Another karyotype pattern occurs in the species of the genus Amazona, and is principally characterized the presence of 8 autosomic macrochromosome pairs with a significant number of telocentric macrochromosomes, including pair 1. Pairs 2, 3 and 4 vary from submetacentric to subtelocentric, and pair 8 is metacentric (De Boer and Belterman, 1980;Van Dongen and De Boer, 1984;Schmutz and Prus, 1987;Aquino and Ferrari, 1990;Lucca et al., 1991;Duarte and Caparroz, 1995).
A further karyotype pattern is observed in the species of the genus Pionus (Lucca et al., 1991) and also in Salvatoria xanthops (Duarte and Caparroz, 1995), and consists between 6 and 8 pairs of macrochromosomes, of which pairs 1, 2, 3 and 4 are submetacentric or subtelocentric.Chromosome pair 5 was described as being subtelocentric in Pionus maximiliani, P. menstruus and P. senilis while in P. seniloides it was described as telocentric (Lucca et al., 1991).However, pairs 5 and 6 are very similar in size in all Pionus species so far studied and it transpires that the P. seniloides pair 5 was erroneously identified as pair 6 (and pair 6 as pair 5), which means that the P. seniloides pair 5 is really subtelocentric and pair 6 telo-  centric, this being true for all Pionus species so far studied as well as for S. xanthops.In addition, Duarte and Caparroz (1995) report the presence of two small metacentric pairs (pairs 7 and 8) in S. xanthops that have been found to be absent in all species of the genus Pionus so far karyotyped.The P. maximiliani karyotype described here is different from the karyotype previously reported for this species and is also different from P. menstruus, P. senilis and P. seniloides (Lucca et al., 1991).all Pionus so far karyotyped, the first four autosome and the Z-chromosome pairs are subtelocentric while in the P. maximiliani specimens analyzed by us they were submetacentric.The karyotypic differences observed between P. maximiliani (previously described), P. menstruus, P. senilis and P. seniloides and P. maximiliani as described by us in this paper could be due to differences in how the arm-ratio was calculated method leading to divergent classification of the same chromosome pair, because of which we consider that these morphological differences are not significant.In addition, the Z-chromosome of P. maximiliani, P. menstruus and P. senilis was reported by Lucca et al. (1991) as being equivalent in size to pair 4, although since only one male of each species was analyzed by Lucca et al. (1991) it appears possible that the largest chromosome pair of these Pionus species correspond to the Z-chromosomes as observed by us in P. maximiliani.
On the other hand, the sexual chromosomes of P. seniloides are very different from those observed by Lucca et al. (1991) in other Pionus and by us in P. maximiliani.Since Lucca et al. (1991) analyzed one female P. seniloides they were able to identify the sexual chromosome pair in this species and found that the Z-chromosome of P. seniloides was subtelocentric and of the same size as pair 4, while the W-chromosome was metacentric and smaller than pair 6.The differences in the sex chromosomes between these species appears to be very significant for understanding the phylogenetic relationships within the genus Pionus (to be discussed below).
The G. brachyurus karyotype described for the first time in this paper is very similar to that previously reported for S. xanthops (Duarte and Caparroz, 1995), except for the diploid number, which is 64 in G. brachyurus and 68 in S. xanthops, and the largest autosome pair which was described as subtelocentric in S. xanthops (Figure 4).This karyotypic similarity between G. brachyurus and S. xanthops suggests that these two species are likely to be closely related, and the similarity between the C-banding patterns (especially for the W-chromosome) observed in these two species also supports this similarity.Both the short arm and the pericentromeric region of the long arm of the W-chromosome are constitutively heterochromatic in G. brachyurus and S. xanthops while in all other New World parrots studied the W-chromosome is largely Cpositive.Very little is known about the patterns of heterochromatin distribution among the Psittaciformes because C-banding information is only available for few species (Mengden, 1981;Lucca, 1983Lucca, , 1984;;Lucca and De Marco, 1983;Aquino, 1987;Aquino and Ferrari, 1990;Christidis et al., 1991), all of which are South American.However, the data available indicate that changes in both the quantity and the distribution of heterochromatin have played a significant role in the chromosomal evolution of this order (Christidis et al., 1991).The close relationship between S. xanthops and G. brachyurus is also consistent with the morphological similarity between these species reported previously (Ribeiro, 1920).Furthermore, according to a recent molecular phylogenetic study involving New World parrots (Russello and Amato, 2003) the genus Amazona is not a monophyletic group, with S. xanthops forming a well-supported sister group relationship with G. brachyurus to the exclusion of all other Amazon parrots.
The karyotypes of G. brachyurus and S. xanthops are more similar to that described here for P. maximiliani and to those reported by Lucca et al. (1991) for other Pionus species than to those reported for other New World parrots.The two metacentric macrochromosomes of G. brachyurus and S. xanthops (pairs 7 and 8), which are absent in all Pionus species, could be a result of a reduction in the diploid number (72 in all Pionus species but 64 in G. brachyurus and 68 in S. xanthops) by Robertsonian translocations between larger microchromosomes.Van Donger and De Boer (1984) pointed out that the small metacentric pairs observed in Melopsittacus undulatus and Ara species were likely to have been formed by Robertsonian translocation between larger microchromosomes, while Christidis et al. (1991) concluded that a karyotype lacking biarmed autosomes is ancestral for the Cacatuidae and that the derived karyotypes were evolved by micro-macrochromosome fusions.
The characteristics of the sexual chromosomes of G. brachyurus and S. xanthops, which are shared by Pionus (except by P. seniloides), are also unique.In all these species, the Z-chromosome is submetacentric and is the largest chromosome in the karyotype while the W-chromosome is submetacentric and equivalent in size to macrochromosome pair 4. The Z-chromosome appears to be conserved in all the remaining New World parrots as it is metacentric and similar in size, corresponding to chromo-some pair 4 or 5.The W-chromosome is morphologically more variable among the New World parrots than the Zchromosome, the W-chromosome having been identified as metacentric in most of the New World parrots, submetacentric in Anodorhynchus hyacinthinus (Lunardi et al., 2003), subtelocentric in Aratinga leucophthalmus (Lucca, 1984) and telocentric in Propyrrhura maracana, Aratinga auricapilla and Deroptyus accipitrinus (Francisco and Galetti, 2001;Lunardi et al., 2003).However, the W-chromosome showed a size similar to the smallest macro-autosomes in all the species studied.
Based on the chromosomal similarity between G. brachyurus, S. xanthops and some species of the genus Pionus, especially for the sexual chromosomes (except for P. seniloides), and the karyotypic differences from the other New World parrots, it appears likely that these species are closely related.In addition, the fact that at least for the New World parrot analyzed so far and belonging to the same genus, such as Amazona, Ara and Aratinga, have very similar karyotypes supports this point of view.Since G. brachyurus and S. xanthops probably share two derived small metacentrics, it seems likely that chromosome rearrangements occurred after divergence from the common ancestor of G. brachyurus and S. xanthops and members of the genus Pionus.This being the case, G. brachyurus and S. xanthops are considered by us to be sister species closely related to the genus Pionus.
As mentioned above, the sexual chromosomes of P. seniloides are different in morphology and size from those described for the other Pionus species and seem more similar to those reported for the New World parrots so far studied and also to some Australian parrots (Van Dongen and De Boer, 1984;Christidis et al., 1991).Based on these data, it seems possible that the chromosome rearrangements responsible for the formation of the karyotypic pattern shared by P. maximiliani, P. menstruus, P. senilis, G. brachyurus and S. xanthops may have occurred after the divergence between the common ancestor of these five latter species and P. seniloides.If this is indeed the case, the genus Pionus, as currently described, is not monophyletic, although the possibility of karyotypic convergence in P. seniloides, while unlikely, cannot be discard.
A close phylogenetic relationship between S. xanthops, G. brachyurus and some Pionus species (P.menstruus and P. senilis) has also been observed in some molecular analyses (Miyaki et al., 1998;Russello and Amato, 2003).However, as the taxonomic sampling of species in these studies was specifically tailored to address the question of monophyly of genera or groups determined a priori, these studies were unable to investigate the species level relationships of the genus Pionus.A forthcoming phylogenetic analysis of all Pionus species is expected to add to our understanding of the phylogenetic relationships within this genus.Finally, to confirm the results presented here, detailed analyses of inter-specific homologies based on chromosome banding methods are needed, both in the species already studied and in many other species not investigated in our study.The present study, and others cited above, demonstrate the usefulness of using chromosomal data to reconstruct evolutionary relationships in the Psittaciformes.Moreover, investigations involving more sophisticated chromosome staining methods, as well as employing other methodologies, such as DNA sequencing, will likely prove useful in fully elucidating the phylogenetic relationships within the Psittaciformes.

Figure 1 -
Figure 1 -Conventionally stained mitotic metaphase and female karyotype showing the macrochromosomes and the first and last pairs of microchromosomes of Pionus maximiliani.Bar = 10 µm.

Figure 2 -
Figure 2 -Conventionally stained mitotic metaphase and female karyotype showing the macrochromosomes and the first and last pairs of microchromosomes of Graydidascalus brachyurus.Bar = 10 µm.

Figure 3
Figure 3 -C-banding patterns of the macrochromosomes of (A) a Graydidascalus brachyurus male and (B) a Salvatoria xanthops female.A G. brachyurus ZW female sex chromosomes is shown in the insert.Bar = 10 µm.

Figure 4 -
Figure 4 -Conventionally stained mitotic metaphase and male karyotype showing the macrochromosomes and the first and last pairs of microchromosomes of Salvatoria xanthops, in conventional staining.In the insert, the ZW female sex chromosomes.Bar = 10 µm.