Chromosome Comparisons of Australian Scaptodrosophila Species

Simple Summary Scaptodrosophila are a diverse group of flies that are widespread in Australia but have received limited research attention. In this study, we characterized the chromosomes of 12 Scaptodrosophila species. We found that the structural changes in their chromosomes are very similar to those seen in Drosophila, a related genus of flies. This includes the amplification of repetitive elements and changes in chromosome length in small (dot) chromosomes and the sex chromosomes. We also found numerous weak points along the arms of polytene chromosomes, which suggest the presence of internal repetitive sequences. Regions producing the nucleolus are at the same chromosome positions in Scaptodrosophila and Drosophila. These chromosomal studies provide a foundation for future genetic studies in Scaptodrosophila flies. Abstract The Scaptodrosophila represent a diverse group of Diptera closely related to Drosophila. Although they have radiated extensively in Australia, they have been the focus of few studies. Here, we characterized the karyotypes of 12 Scaptodrosophila species from several species groups and showed that they have undergone similar types of karyotypic change to those seen in Drosophila. This includes heterochromatin amplification involved in length changes of the sex and ‘dot’ chromosomes as well as the autosomes, particularly in the coracina group of species. Numerous weak points along the arms of the polytene chromosomes suggest the presence of internal repetitive sequence DNA, but these regions did not C-band in mitotic chromosomes, and their analysis will depend on DNA sequencing. The nucleolar organizing regions (NORs) are at the same chromosome positions in Scaptodrosophila as in Drosophila, and the various mechanisms responsible for changing arm configurations also appear to be the same. These chromosomal studies provide a complementary resource to other investigations of this group, with several species currently being sequenced.


Introduction
The genus Scaptodrosophila is estimated to have diverged within the drosophilid lineage during the Cretaceous period about 70 million years ago. Further diversification occurred from the Eocene through the Holocene, about 50-0.7 million years ago [1]. The Scaptodrosophila were originally considered a subgenus of Drosophila and called Pholadoris in earlier studies [2,3], but in a taxonomic revision of the family, Grimaldi [4] elevated the group to genus rank, which was supported by molecular data from DeSalle [5].
Of the 36 drosophilid genera represented in Australia, approximately a third of the described species belong to the genus Scaptodrosophila. This prevalence of Scaptodrosophila in the Australasian drosophilid radiation is unique when compared to diversification elsewhere [6,7]. However, the genus has been the subject of relatively few scientific studies [8,9]. The developmental requirements of most Scaptodrosophila species are unknown and appear to be restrictive since few species are widespread in urban environments [6]. Scaptodrosophila species have been reported as feeding or breeding on tree sap, fungi, fruit and flowers [8,10]. Unlike Drosophila species, most cannot be collected on fruit baits, except for members of the coracina species group [11]. Scaptodrosophila hibisci and S. aclinata breed in the flowers of native Hibiscus species [7,12]. There are also species that feed on and form galls in the stems of a bracken fern [13], while others feed on eucalyptus leaf litter in the crowns of tree ferns [8]. Species from the genus may have important ecological functions such as in pollination and nutrient cycling, as well as in acting as a food source for other organisms.
Some coracina group species can be reared on conventional Drosophila laboratory medium [9]. However, other species must be supplied with fruit or flowers to stimulate egg laying. In the laboratory, mature larvae usually complete development by leaving their food vial and burrowing into surrounding sand or vermiculite to pupate [8]. Their 'skipping' or springing behavior facilitates this and likely aids their dispersal in nature. The practice of burrowing into ground material to pupate may have evolved as a mechanism to protect the quiescent stage from environmental extremes and fires as well as predation and parasitism. It is not surprising, therefore, that most species have been difficult to maintain in the laboratory, particularly in long-term culture [8].
Given the difficulties in capturing and maintaining Scaptodrosophila species, few chromosomal studies have focused on Scaptodrosophila, with the most complete study carried out by Bock [11] on six members of the lativittata species complex of the coracina group. An early study by Mather [3] presented a scheme of chromosome evolution. However, he considered them a subgenus of Drosophila, and his diagram of chromosome evolution mixed Drosophila and Scaptodrosophila (Pholadoris) species. A more recent study by Wilson et al. [12] on S. hibisci indicated the presence of male recombination and neo-sex chromosome formation in this species.
Here, we established conditions for the short-term culture of several Scaptodrosophila and used these lines to karyotype some species. Since Scaptodrosophila and Drosophila species have been separated for about 70 myr and adult morphology is still quite similar, we tested if the karyotypic changes that had occurred in species from these groups were also similar. The cerebral ganglion chromosomes of the species that we managed to culture were examined morphologically and also subjected to C-banding for constitutive heterochromatin localization and fluorescence in situ hybridization (FISH), to locate ribosomal RNA genes. We attempted silver staining to examine the location of active nucleolar organizing regions (NORs). We examined the polytene chromosomes but did not attempt a detailed banding comparison.

Species Examined
Species were mainly collected in northeastern Australia and identified by one of us (M.S.) ( Table 1). Several species are either new records for Australia or undescribed, and their classification is currently under investigation.
Isofemale lines were established in the laboratory for the collected species. Although most of the Scaptodrosophila studied here have white larvae, larvae of S. novoguineensis and S. sp. aff. novoguineensis become purple as they age. This purple pigment occurs in both the fat body and hemolymph but appears to originate from the fat body. All feeding larvae had long posterior spiracles and showed the characteristic skipping behavior. The species that we studied preferred pupating in vermiculite that surrounded the culture vials. However, we cultured two species, S. sp. aff. concolor strain CBN17 and S. lativittata, for a period solely in food vials, and S. xanthorrhoeae can also be maintained indefinitely without access to a vermiculite substrate.
Polytene chromosomes of S. nitidithorax were examined for inversion polymorphism. The mitotic karyotype of this species comes from Bock [11]. Mitotic and polytene chromosomes of a small number of S. specensis had also previously been examined by Bock [11]. We C-banded several ganglion chromosome preparations of S. specensis. Polytene and mitotic chromosomes were examined for S. bryani and S. sp. aff. concolor strain CBR1. These techniques along with C-banding were completed for S. novoguineensis, S. lativittata and S. evanescens. All methodologies plus NOR localization were completed for S. xanthorrhoeae, S. claytoni, S. cancellata, Scaptodrosophila sp. aff. cancellata, S. sp. aff. concolor strain CBN17 and Scaptodrosophila sp. aff. novoguineensis. The chromosome techniques used were the following: squash preparations of salivary gland polytene chromosomes in lacto-aceto orcein, squashes or drop spreads of brain mitoses stained with lacto-aceto orcein or Giemsa (Gibco, Adelaide, Australia), C-banded spreads of brain mitoses stained with Giemsa, NOR silver staining of mitotic spreads and fluorescence in situ hybridization (FISH) to identify the NORs (Table 2).

Polytene Chromosome Squashes
Glands were removed in Drosophila Ringer's solution and fixed in 50% acetic acid for 7-10 s and 1N HCl for about 30 s. They were then moved to a small drop of mounting medium (60% acetic acid: lactic acid, 1:1) on a siliconized coverslip and allowed to clear before a very small amount of lacto-acetic orcein stain was added and the coverslip picked up with a cleaned slide and tapped/squashed to spread the chromosomes. To visualize the nucleolus, some preparations were fixed in 3:1 ethanol:acetic acid or methanol:acetic acid.

Cerebral Ganglion Chromosome Spreads
The cerebral ganglions were removed in Drosophila Ringer's solution, placed in 0.5% Na citrate for 10-15 min to spread the chromosomes and then fixed in several changes of methanol:acetic acid:water (11:11:1) for 1-2 min or longer. Tissue was then stained in lacto-acetic orcein on a siliconized coverslip for 10 min, a drop of 60% acetic acid was added and the preparation was picked up with a cleaned slide and squashed. For slides to be used in C-banding, NOR silver staining or FISH, tissue was either squashed in 60% acetic acid and the coverslip removed after freezing in liquid nitrogen, or the cells were allowed to disassociate in 60% acetic acid and small volumes picked up with a micro-pipette, dropped on a clean, warmed slide and air dried.

C-Banding
C-banding generally followed the technique used by Bedo [14] with a few variations. We used a 2-3 min incubation in 5% Ba(OH) 2 at 50 • C followed by extensive washing in running tap water and deionized water. The Ba(OH) 2 did not go completely into the solution, and extensive washing in warm tap water was important for getting rid of some of the precipitate. This was followed by air drying, incubation for 1 h in 2X SSC (saline sodium citrate) buffer at 60 • C and a rinse in deionized water. The slide was then either stained immediately in 4% Giemsa (Gibco) for 15-30 min, or dried and stained the following day.
This was followed by a quick rinse in deionized water. After drying well, the preparation was mounted in Gurr's Neutral Mounting Media and examined.

NOR Silver Staining
NOR silver staining was based on the technique of Howell and Black [15]. Since staining on mitotic chromosome figures was not consistent, various modifications of temperature and time [16][17][18], pH [19], pre-and post-treatment of slides [20][21][22][23][24] and age of the slide were tested. None were satisfactory in terms of giving consistent results. We tried heating in a microwave oven [25], but our preparations tended to be unevenly developed with no consistent deposition of silver grains over a particular chromosome.

Fluorescence In Situ Hybridization (FISH)
Fluorescence in situ hybridization was carried out using two methods. The same result was achieved by both. 1. A Drosophila melanogaster DNA clone, pDm238 in pBR322 [26], was labeled by nick translation and used as a probe as described previously [27]. This probe was an 11.5 kb fragment that contained the 18S, 5.8S and 28S genes of D. melanogaster plus an intergenic spacer. 2. Total Drosophila RNA was used as a probe following the technique of Madalena et al. [28] or a slightly modified version of this technique. Approximately 40 Drosophila melanogaster flies were etherized and immediately frozen in liquid nitrogen. They were homogenized in 0.8 mL Trizon (Invitrogen, Adelaide, Australia), and chloroform was added to a total volume of 1 mL with mixing by Eppendorf inversion. After 15 min at 4 • C, the preparation was centrifuged at 12,000 rpm at 10 • C for 10 min. One volume of frozen isopropanol was added to the extracted liquid phase with mixing, and the Eppendorf was kept in the freezer for 5 h. To recover the RNA, centrifugation was repeated as above and the pellet resuspended in formamide containing poly-(r)U (Sigma, Adelaide, Australia) (1 µg/mL).
Chromosome spreads were treated with RNase A diluted in 2X SSC (0.2 mg/mL) for 3 h at room temperature. The slides were washed in 0.5X SSC, and chromosome DNA denaturation was performed by one of two techniques. In the first, slides were treated as in Madalena et al. [28]. A probe mixture (6 µL) of 50% formamide, 2X SSPE (saline sodium phosphate EDTA buffer), 0.1% SDS and 100 ng of insect RNA mixed with poly (r)-U was applied to each air-dried slide and covered with a plastic coverslip. The slides were steam heated at 75 • C for denaturation and then kept in a closed plastic box at 37 • C overnight or longer for hybridization. In the second technique, slides were incubated in 0.08 N NaOH for 1 min at room temperature and then washed in 0.5X SSC and kept in ice-cold ethanol until the hybridization procedure was carried out. The slides to be hybridized were dried, and the hybridization mixture, consisting of 50% formamide containing RNA and 3X SSC, was applied under a plastic coverslip. The slides were incubated overnight at 37 • C.
Detection of the DNA-labeled probe was described in Stocker et al. [27]. The RNA probe was detected by incubation with goat IgG anti-RNA.DNA hybrid [29] and detected with rabbit IgG anti-goat labeled with TRITC (Sigma), as described in Madalena et al. [28] and Gorab et al. [30]. Slides were counterstained with 4 , 6-diamino-2-phenylindole (DAPI). They were mounted in Vectashield anti-fading solution (Vector Labs, Adelaide, Australia) and inspected with epifluorescence optics using Zeiss and Olympus microscopes equipped with CCD cameras and image analysis software.

Lacto-Acetic Orcein Staining
The mitotic karyotypes of the 12 Scaptodrosophila species are shown in Table 3. We have also included karyotypes of the other coracina species whose chromosomes were examined by Bock [11] as well as those of S. hibisci studied by Wilson et al. [12]. Most of the species examined are members of the coracina group. Two have tentatively been placed in the barkeri group, one is in the bryani group and the rest are not currently grouped with other species. Scaptodrosophila novoguineensis and S. sp. aff. novoguineensis are different species but very similar morphologically. Table 3. Karyotypes of the species (M = large metacentric; m = small metacentric; J = submetacentric; R = rod or acrocentric; D = dot) * from Bock [11], and # from Wilson et al. [12]. The karyotypes of nine species in the coracina group are very similar, particularly with respect to their autosomes. The tenth species, S. sp aff. cancellata, currently undescribed, lacks the metacentric autosome and instead has two acrocentrics (rods). The major difference among these karyotypes is in the shape of the 'dot' and sex chromosomes. 'Dot' chromosomes range from very small 'dots' (claytoni, evanescens, specensis) to large metacentrics (enigma). The X and Y chromosomes also have variable shapes from acrocentrics to submetacentrics.

Species
Scaptodrosophila sp. aff. concolor strain CBN17 has a superficially similar karyotype to the species in the coracina group, with a submetacentric X chromosome and a somewhat elongated 'dot' chromosome (Table 3, Figure 1A insert). The Y mainly appears as an acrocentric. A 'dot' chromosome is clearly apparent in the polytene set and could often be distinguished by its proximity to the nucleolus ( Figure 1A). Polytene chromosomes in the S. sp. aff. concolor strain CBN17 lines that were examined are also polymorphic for a large terminal inversion on one of the autosomes ( Figure 1A,B). The other end of this autosome shows pairing between repetitive sequences which is common in Scaptodrosophila ( Figure 1B). Two additional polymorphic inversions were observed in the original isofemale line of S. sp. aff. concolor strain CBN17. The other species that we studied showed a variable number of arm fusions. Scaptodrosophila sp. aff. concolor strain CBR1 has two pairs of metacentric autosomes and one pair of acrocentric autosomes. The 'dot' and X chromosomes have short arms, and the Y often shows a central constriction (Figure 2A). Scaptodrosophila sp. aff. concolor strain CBN17 and S. sp. aff. concolor strain CBR1 have been tentatively placed in the barkeri group taxonomically, due to their superficial similarity to S. concolor, but this is likely to be revised upon examination of these species compared with type specimens of S. concolor, particularly in light of their chromosomal differences. The chromosomes of S. sp. aff. concolor strain CBR1 more closely resemble those of S. xanthorrhoeae ( Figure 2B), which has not been assigned to a group. However, S. xanthorrhoeae has an acrocentric X and Y and small 'dots'. Scaptodrosophila novoguineensis and S. sp. aff. novoguineensis show further reductions, having two metacentric autosome pairs, an acrocentric-shaped X and Y and small 'dots'. Scaptodrosophila bryani has the most numerically reduced karyotype with two metacentric chromosome pairs, an acrocentric X and Y and no 'dot'. Scaptodrosophila hibisci [12] has a complex karyotype that appears very different from the Scaptodrosophila karyotypes that we have examined here. It has a large metacentric chromosome pair with variable Giemsa staining of homologues, three small metacentric pairs, a small acrocentric and a small 'dot' chromosome pair. Experiments by Wilson et al. suggested neo-sex chromosome formation in this species [12]. While sex chromosomes have not been definitively identified, the authors suggest that the large metacentrics could represent sex chromosomes, although additional translocations could have brought other chromosomes into the sex-determining system [12].
The quality of the polytene spreads in the species we examined varied, and good spreads proved difficult to obtain for most species. Scaptodrosophila chromosomes have numerous 'weak points' which are prone to breakage and the extrusion of small chromosome segments. This made it difficult to define individual chromosomes and distinguish the 'dot' chromosome from small segments of other chromosomes that had become detached. An extreme example is shown in the squash of an S. bryani nucleus ( Figure 3A). Besides S. sp. aff. concolor strain CBN17, possible dots could be observed in polytene sets of S. cancellata ( Figure 3B) and S. sp. aff. cancellata which both had metacentric dot chromosomes in mitotic spreads. 'Dot' chromosomes in polytene sets of other species were not repeatedly observed.

C-Banding
C-banding was carried out to examine the role of heterochromatin addition in karyotype changes among these Scaptodrosophila species. Some of the species examined by Bock [11] had large 'dot' chromosomes. Enlarged 'dot' chromosomes were also observed for several of the species in our investigation. We would expect these chromosomes to be C-band-positive. We were also interested in examining other chromosomes for the addition of C-banded heterochromatin regions. C-banding of ganglion cell chromosomes of the six coracina group species that were available gave the results shown in Figure 4. The three species with very small 'dot' chromosomes, S. claytoni, S. specensis and S. evanescens, have a similar C-banding pattern ( Figure 4A). C-banded heterochromatin is consistently located at either side of the centromere of the metacentric autosomes and at the centromeric end of the acrocentric autosomes and the X chromosome. The Y chromosome is entirely C-band-positive. Occasionally, internal C-bands were observed, such as those on the X chromosome of S. claytoni, but these were not always present. Surprisingly, the 'dot' chromosomes of these three species are not always C-band-positive. Three species have larger, metacentric dot chromosomes. Except for its metacentric, C-band-positive 'dot' chromosome, S. lativittata has a C-banding pattern similar to that described for the species above. Scaptodrosophila cancellata has more C-banding heterochromatin at the centromeric ends of its acrocentrics than the other species. Both S. cancellata and S. sp. aff. cancellata have submetacentric X chromosomes with C-band-positive short arms and prominent Cband-positive metacentric 'dot' chromosomes ( Figure 4B,C). Scaptodrosophila sp. aff. concolor strain CBN17 ( Figure 4D) is not a member of the coracina group but, like them, has a pair of metacentric autosomes with C-banding near the centromere and C-bands near the centromeric end of the three acrocentric autosomes. However, the 'dot' chromosomes of this species have one part somewhat more darkly stained than the other, and the submetacentric Xs have only a fine C-band at the region of the presumed centromere ( Figure 4D).  Scaptodrosophila xanthorrhoeae and S. sp. aff. concolor strain CBR1 have similar karyotypes (Table 3), but S. sp. aff. concolor strain CBR1 was lost before C-banding had been carried out. The C-banded karyotype of S. xanthorrhoeae is shown in Figure 4E. C-banded heterochromatin is observed at either side of the centromere of the metacentric chromosomes and at the centromeric ends of the acrocentrics. The 'dot' chromosome is C-bandpositive in many, but not all, sets. Scaptodrosophila sp. aff. concolor strain CBR1 ( Figure 3A) has a small, metacentric 'dot' chromosome which often stains darkly in orcein squashes, similar to the long, apparently metacentric Y chromosome. This likely indicates that they would be C-band-positive.
Scaptodrosophila novoguineensis has the same karyotype as S. sp. aff. novoguineensis. C-banding is shown for S. sp. aff. novoguineensis, a new species which is currently being described ( Figure 4F). C-band-positive heterochromatin is located at either side of the centromere of the two metacentric chromosomes and at the centromeric end of the X chromosome. The Y chromosome is totally C-band-positive. The dot chromosome is positive in this preparation.
In these Scaptodrosophila species, no consistent internal C-bands in mitotic chromosomes were detected with the technique used, although the number of weak points and inverted repeats in polytene chromosomes suggested the presence of internal repetitive DNA.
C-banding was conducted by Wilson et al. [12] for cerebral ganglion chromosomes of S. hibisci and is shown for one of the species' two chromosome 1 forms [12]. Considerable C-banded heterochromatin is located on all chromosomes, including entire arms of chromosomes 2 and 3 and almost an entire arm of chromosome 1. It is clear that heterochromatin additions and deletions have been important in shaping the karyotype of this species, perhaps changing acrocentrics to metacentrics, but further study is required to decipher the changes that have occurred.

NOR Location
The location of the NOR was examined in S. cancellata, S. sp. aff. cancellata, S. claytoni, S. sp. aff. concolor strain CBN17, S. xanthorrhoeae and S. sp. aff. novoguineensis. For all species except S. sp. aff. concolor strain CBN17, in situ hybridization located the ribosomal DNA genes on the sex chromosomes. In S. xanthorrhoeae ( Figure 5A,B), S. claytoni and S. sp aff. novoguineensis, one NOR was at the very end of the short heterochromatic block on the acrocentric X, probably just above the heterochromatin. Scaptodrosophila cancellata and S. sp. aff. cancellata had a submetacentric X, with the short arm almost completely heterochromatic ( Figure 4B,C). In both species, the NOR was at the end of this arm ( Figure 5C-F). In all of these species except S. claytoni, we found another NOR at one end of the Y chromosome ( Figure 5B,D,F). For S. claytoni, we could not ascertain if the Y chromosome contained an rDNA site because we never managed to find a male included among our in situ slides. For S. lativittata, which was not considered in the in situ hybridizations, the presence of distinctive satellites at the heterochromatic end of the X chromosomes indicated that this was one of the NOR sites (A. Stocker, personal observations). In several species, there appeared to be variation in the number of ribosomal cistrons between the two X chromosomes, but this needs further investigation.
Scaptodrosophila sp. aff. concolor strain CBN17 differed from the other species in having its ribosomal genes on the dot chromosome ( Figure 6A-C). There was very little heterochromatin present on the X chromosome of this species (Figure 4D), and neither the X nor Y showed any rDNA. Attempts to identify the site of the active nucleolus in mitotic spreads using variations of the silver staining technique of Howell and Black [15] were mostly unsuccessful. The only consistent silver staining was at the presumed nucleoli of interphase cells ( Figure 7A1). Occasionally, a silver-stained nucleolus or fragment of a nucleolus could be seen attached to a prophase chromosome, but it was usually unclear which chromosome was involved ( Figure 7A1). When staining occurred on metaphase chromosomes, it was usually at centromeres. Rarely, staining occurred at a chromosome region that we knew from FISH results contained rDNA ( Figure 7B). Most mitoses did not show silver staining at all. Such variable results have also been observed in other species [22,31,32].

C-Banding
C-banding identifies constitutive heterochromatin blocks which are more resistant than euchromatin to DNA removal [33]. The C-banding pattern of the Scaptodrosophila chromosomes was generally consistent (Figure 8), with the exception of the smaller 'dot' chromosomes which sometimes did not C-band. The heterochromatic nature of the 'dot' chromosomes is a conserved feature throughout the drosophilids. However, their repeat level is considerably lower than pericentric heterochromatin since other gene sequences are dispersed within them [34]. Since the genes in 'dot' chromosomes of different species are apparently the same, smaller 'dot' chromosomes should contain fewer repeats than larger 'dots' and therefore be less heterochromatic. This could at least partially explain their lack, or variability, of C-banding in this study. In all the species we studied, the Y chromosome C-banded very strongly. It was usually impossible to discern clear structural variation along this chromosome, although sometimes there were constrictions or a region that looked darker than the rest. If the strength of C-banding is related to a high repeat level, the Y chromosome DNA must be very highly repetitious. There were no consistent C-bands within the arms of the mitotic chromosomes despite the presence of numerous weak points, indicating internal heterochromatin along the polytene arms.
The extensive C-banded regions in S. hibisci [12] suggest that the distinctive karyotype of this species may have been obtained by heterochromatin addition to acrocentrics, changing them to metacentrics or submetacentrics. The authors suggest that the rearranged banding between the chromosome 1 homologues is probably caused by a pericentric inversion. Chromosome 1 is seen in four forms and two combinations. The forms with the end satellite have characteristics of the X chromosome, with the satellite indicating the position of the NOR, which is commonly found on the X in drosophilids. The thin, intermediately banded chromosome that it is paired with in Type 1 cells is suggested by the authors to be a partially heterochromatized neo-Y chromosome. C-banding is not shown for this chromosome 1 homologue, but if the pale Giemsa staining corresponds to the absence of C-bands, the weak staining of this homologue may mean that other factors in addition to type and number of repeats are responsible for the presence of C-banding. Translocations between crucial Y regions and other chromosomes may also have played a role in the staining differences.

NOR Location
Most species in the genus Drosophila that have been examined thus far have NORs on their sex chromosomes, and this is believed to be the ancestral condition [35]. NORassociated sequences probably assist in pairing between the sex chromosomes during meiosis [36,37]. On the Drosophila X chromosomes, NORs were usually located near the centromere at heterochromatic regions [35]. This was also the case in three of the Scaptodrosophila species for which the NOR position was examined ( Figure 8). In Drosophila hydei, however, the X is metacentric and the NOR is located near the distal end of its heterochromatic arm [38]. This position is analogous with the results we obtained in S. cancellata and S. sp. aff. cancellata ( Figure 8) and suggests that heterochromatin amplification between the centromere and the NOR moved this region toward the end of the arm during species evolution. The Y chromosome is entirely heterochromatic in Drosophila species, and NORs on this chromosome are more widely distributed [35]. In the group of Scaptodrosophila species examined, the NOR was always observed at one end of the Y (Figure 8). Some species in the genus Drosophila have NORs on chromosomes other than the sex chromosomes. Most of the members of the D. ananassae group have a NOR site on the 'dot' chromosome [35]. One of the Scaptodrosophila species that we examined, S. sp. aff. concolor strain CBN17, also has its NOR on the dot chromosome ( Figure 8). Two Hawaiian Drosophila species have a NOR on a different, non-heterochromatic autosome (the B element according to Mueller's [39] nomenclature), but the band at which it is located is heterochromatic [40]. Despite the changes in the NOR position, it is consistently associated with heterochromatin regions.

Karyotype Evolution
In Drosophila, karyotype examination [39,41] and more recent sequencing [42] have given convincing evidence that the primitive chromosome configuration is five pairs of acrocentrics (rods), including sex chromosomes, and one pair of 'dots', which are usually very small and heterochromatic. When the chromosome number differed from this, gene comparisons usually could identify six representative 'elements' called A-F, with A being the X chromosome and F the 'dot' chromosome. The Scaptodrosophila species in the current study have chromosome numbers ranging from five pairs of acrocentrics, a submetacentric X chromosome pair and a pair of 'dots', to two metacentric pairs, acrocentric sex chromosomes and no 'dots'. Using arm changes and heterochromatin variation, we can suggest how some of these karyotypes originated from a hypothetical primitive one.
The most common karyotype among the species that we have examined is one pair of metacentrics and three pairs of acrocentric autosomes, acrocentric or submetacentric sex chromosomes and 'dot' chromosomes of varying sizes. This karyotype is found in all except one member of the coracina group species and also in S. sp. aff. concolor strain CBN17. It can be derived from the primitive karyotype by postulating a pericentric inversion in an acrocentric autosome that moved the centromere to a central position, pericentric inversions and heterochromatin amplification in some of the sex chromosomes and heterochromatin amplification in 'dot' chromosomes ( Figure 9). The exceptional karyotype among the coracina species is S. sp. aff. cancellata, which has five pairs of acrocentric autosomes, submetacentric sex chromosomes and large metacentric 'dot' chromosomes. This karyotype could have been obtained by dissociation of the metacentric autosome into two acrocentrics. Heterochromatin changes are very similar to those in S. cancellata. Since other members of the coracina group have an acrocentric X [11], and in S. claytoni, the NOR is associated with centromeric heterochromatin, it is most probable that the heterochromatic arm of the X developed through extension of heterochromatin between the centromere and the NOR, leaving the NOR positioned near its end. Changes in the length of 'dot' chromosomes were common in the coracina species examined by Bock [11] (Figures 1-6). Scaptodrosophila sp. aff. concolor strain CBN17 was tentatively placed in the barkeri species group because of its superficial similarity to S. concolor, but its karyotype looked similar to that of the coracina species until the NOR sites were located. These were found to be on the 'dot' chromosomes rather than on the X and Y (Figure 10). The yellow/brown bodies and reddish/orange eyes of the adult flies were also quite different from the darker color of the coracina species that we examined. The changes that formed the S. sp. aff. concolor strain CBN17 karyotype are hypothesized to have been the same pericentric inversion described above, and transposition of the ribosomal DNA cistrons from the X to the dot chromosome as a pericentric inversion was formed in the X (Figure 10). In this process, the X appears to have lost most of its heterochromatin. It is possible that the rDNA on the dot chromosome came from the Y, since Y rDNA cistrons are also absent. To resolve this question, we would need an intermediate situation with a NOR on the dot chromosome and another on one of the sex chromosomes. With respect to its karyotype, this species appears to have diverged directly from the ancestor of the coracina group ( Figure 10). The other species show a reduction in chromosome number and changes in chromosome shape, with elongation of chromosome short arms in S. sp. aff. concolor strain CBR1, also tentatively placed in the barkeri species group (Figure 11). However, S. sp. aff. concolor strain CBR1 looks chromosomally similar to S. xanthorrhoeae (Figure 11), with a reduction in acrocentric autosomes to one and the gain of a smaller metacentric. A pericentric inversion could have formed the smaller metacentric, and a centromeric fusion between the two acrocentrics would form the larger acrocentric. Although we do not have C-banding for S. sp. aff. concolor strain CBR1, its short arms, not present in S. xanthorrhoeae, and larger metacentric 'dot' chromosomes are probably heterochromatic. Further arm reduction is observed in S. novoguineensis and S. sp. aff. novoguineensis, both having karyotypes composed of two pairs of metacentrics, a pair of dots and an acrocentric X chromosome. The most straightforward manner that these changes could have evolved from a 5R, 1D karyotype is for two centric fusions to have occurred between four pairs of autosomes ( Figure 11). Finally, S. bryani has lost the 'dot' chromosome, most likely by fusion with one of the longer chromosome arms. The absence of a 'dot' chromosome has been observed in several Drosophila species groups. In D. busckii [43], a basal Drosophila species, 'dot' chromosome genes are located near the heterochromatin of the X chromosome and also at a homologous region of the Y chromosome [44]. In situ hybridization also placed 'dot' chromosome genes of D. lebanonensis near the centromere of the acrocentric X chromosome (A element) [45]. Since X and 'dot' chromosomes have some similar characteristics, it has been suggested that, in the ancient history of drosophilids, the 'dot' chromosome (F element) was part of the X chromosome [34]. In fact, sequencing and mapping of non-Drosophila Diptera have shown that the Mueller A element is an autosome and the 'dot' chromosome is the female sex chromosome in these outgroup species [46]. Based on these results, we might hypothesize that 'dot' genes would be translocated near the centromere of the X chromosome in S. bryani. However, in six species of the willistoni species group, 'dot' chromosome genes were located near the centromere of the E element, also acrocentric [47]. Since the X and Y are the only acrocentrics in S. bryani, it seems likely that the F element would be translocated to the X chromosome in this species. However, confirmation of this and other arm translocations will only come from future in situ hybridizations.

Conclusions
The Scaptodrosophila species we have examined indicate that, although this genus separated within the drosophilid lineage relatively early, the types of karyotypic change are similar to those observed in Drosophila species and other features are also shared (Table 4). Heterochromatin amplification appears to have been an important mechanism for structurally changing the arm length of the sex and 'dot' chromosomes in both Scaptodrosophila and Drosophila. Such amplification apparently adds arms to other chromosomes as well [12]. In Drosophila species, heterochromatin amplification seems to have been concentrated in specific groups such as members of the ananassae species subgroup [35] and several Hawaiian groups [48] where the mechanism and basis of this process have been discussed. In Scaptodrosophila, the amplification seems to be particularly important in the coracina species group. However, not enough species have been examined by modern banding techniques to be definitive about this. In the Drosophila karyotype list compiled by Clayton [49], a number of Scaptodrosophila karyotypes do not show a 'dot' (F) chromosome. One reason for this could be the presence in some species of a large 'dot' that was not recognized as heterochromatic by the techniques used. Heterochromatin amplification could be a widespread mechanism of karyotype evolution in the Scaptodrosophila. It is present in S. sp. aff. concolor strain CBR1, potentially a member of the barkeri group, and in S. hibisci, one of the flowerbreeding species, where it could be related to neo-Y formation and male recombination. The locations of the NOR are the same in both Scaptodrosophila and Drosophila [35], as are the various mechanisms responsible for changing arm configurations. The many weak points and apparent inverted repeats observed in polytene chromosomes of Scaptodrosophila species may indicate a higher number of transposable elements in this group, but proof of this would require more investigation at the DNA level. Of the estimated 230 described species of Scaptodrosophila [50], relatively few have been the subject of genetic studies. Preliminary whole genome DNA sequencing of some of these species (Rahul Rane, personal communication) has placed S. sp. aff. concolor strain CBN17 as a relatively recent divergence from the line leading to coracina species in the Scaptodrosophila tree. S. sp. aff. concolor CBR1 was not examined in these DNA comparisons, but the difference between the chromosomes of the two species suggests that, although the flies may look superficially similar to S. concolor, they may not be closely related and could actually belong to different species groups. The line leading to S. xanthorrhoeae diverges earlier in the DNA sequence comparisons, while S. bryani and S. novoguineensis form a separate and even earlier branching. Drosophila lebanonensis has been removed from the Scaptodrosophila lineage. The current analysis should provide a cytological framework for these ongoing genomic efforts.