Chromosomal Phylogeny of the Drosophila Fasciola Species Subgroup Revisited (diptera, Drosophilidae)

The analysis of polytene chromosomes in 26 strains of seven species in the Drosophila fasciola subgroup, from several locations in Brazil, in addition to strains of two species belonging to the Drosophila mulleri subgroup (D. aldrichi and D. mulleri), enabled us to determine that the 3c inversion found in the latter species differ in one of its break points from that present in the species of the fasciola subgroup. Therefore, a change in the mulleri complex denomination from inversion 3c to inversion 3u is proposed. Accordingly, the fasciola subgroup is no longer a lesser phylogenetic part within the mulleri subgroup. Rather, it is directly related to the likely ancestor of the repleta group, called Primitive I. This information removes the main obstacle to considering the Drosophila fasciola subgroup as an ancestral group within the Drosophila repleta species group, according to the hypothesis of Throckmorton. Our data also support the conclusion that D. onca and D. carolinae are closely related species based on one new inversion in chromosome 4 (4f 2), in both species. D. fascioloides and D. ellisoni also form a pair of sister species based on the presence of fusions of chromosomes 2-4 and 3-5. D. rosinae is related only to the likely ancestor of the fasciola subgroup, where the 3c inversion was fixed.

Moreover, in open vegetation, D. rosinae emerged from columnar cacti (Cereus sp.) (Wasserman, 1962a;Pereira et al., 1983;Tidon-Sklorz and Sene, 1995c).Apparently, the morphology of the testicles and of the seminal receptacle of the species in this subgroup is intermediate between that in the mulleri and repleta subgroups (Wasserman, 1962a).
The origin and adaptive radiation of the cactophilic species in the repleta group probably occurred in the Oligocene and Miocene (Throckmorton, 1975(Throckmorton, , 1982)).This group is likely to have originated in the transition zone between the Nearctic and Neotropical biogeographic regions in Mexico (Wasserman, 1954).
Cytologically, the ancestral species of all fasciola subgroup would have evolved from the Primitive I, a hypo-thetical sequence of polytene chromosomal bands, suggested by Wharton (1942), that differs from the standard arrangement of the Drosophila repleta by the presence of the Xabc;2ab;3b inversions, and by the fixation of the 2o 3 , 2e 3 and 2l 3 inversions.Accordingly, the basic chromosomal composition of the subgroup would be the Primitive VII (Xabc; 2abo 2 e 3 l 3 ; 3b; Wasserman 1960Wasserman , 1962aWasserman , 1992)).
According to Wasserman (1982), the existence of the 3c inversion in species of the mulleri complex (included in the mulleri subgroup) and of the fasciola subgroup sustains the hypothesis of a common ancestry, with the fasciola subgroup stemming from the mulleri subgroup.The fact that the species of the fasciola subgroup are forest dwellers could be an indicator of reinvasion of the forests by desertadapted species.Throckmorton (1982), and in earlier works (Throckmorton 1962(Throckmorton , 1975)), discusses the problem of the origin of the repleta group and states: "whether the ancestor of the repleta group itself was a forest species which "became" a "repleta", began diversifying in the forest, and subsequently moved into arid habitats, or whether it first moved into arid habitats and became a repleta there, is difficult to determine.Its closest relatives, the castanea, canalinea, dreyfusi and mesophragmatica groups, are forest forms, for the most part, and apparently primitive members of the repleta group that are at least facultative forest forms breeding in fallen fruit.Parsimoniously, this permits the inference that the founder of the repleta group was a forest form, not necessarily of the wet forest, which "became" a repleta while still associated with forest habitats.At the present time, and on anatomical grounds especially, the major separation within the repleta group is between the hydei subgroup on the one hand and the remaining subgroups on the other, with the fasciola subgroup being the most primitive among the latter forms".Supporting this idea, Morais et al. (1995) proposed the possibility of the repleta group ancestor having inhabited the forests, and, based in composition studies of yeasts, associated to these flies, suggesting that the fasciola subgroup represents the oldest lineage from which the South American species of the repleta group may have evolved.This statement is in agreement to the ecological data mentioned above.
Even though Throckmorton's hypothesis (1962Throckmorton's hypothesis ( , 1975Throckmorton's hypothesis ( and 1982) ) rests upon morphologic and ecological data, it does not explain the presence of the 3c inversion in both fasciola and mulleri subgroups, favoring the hypothesis proposed by Wasserman (1962a and b;1963;1982).

Material and Methods
We analyzed 26 isofemale strains of seven species in the fasciola subgroup established from specimens collected (Tidon and Sene, 1988;Tidon-Sklorz and Sene, 1992) in different locations (Table 1).

562
Diniz and Sene Polytene chromosomes from the salivary glands of third instar larvae were prepared by squashing techniques in 2% lacto-aceto-orcein, fixed in acetic acid and perchloric acid.They were then compared with the maps depicted by Wharton (1942) and Wasserman (1962a).
In order to define the presence of the 3c inversion, the polytene chromosomes of two species of Drosophila belonging to the mulleri complex of the mulleri subgroup (D. aldrichi and D. mulleri) were also analyzed.

Results
Of the seven Drosophila species in the fasciola subgroup under study, five came from humid coastal and inland Brazilian forests, one from Panamanian forests and the remaining one from the Caatinga domain (Table 1).
By comparing chromosome 3 of species in the mulleri (D. aldrichi and D. mulleri) and fasciola subgroups, we verified that one of the breakpoints of the 3c inversion present in the species of the fasciola subgroup is not the same as determined for the 3c inversion present in the species of the mulleri group.That is, there are two overlapped inversions sharing one breakpoint, and not one single inversion as previously thought.The 3c inversion was described by Wasserman (1962a), by analyzing species from the fasciola subgroup, as having the E4a and G1c breakpoints.However, we observed that the inversion present in the species of the mulleri subgroup differs by one of the two break points: E5b -G1c.Thus, we suggest that this inversion in the mulleri subgroup, as it is still undescribed, should be renamed as 3u.(Figures 1 and 2; Table 1).In addition, two new inversions fixed in chromosome 4 were observed along with the inversions reported in the literature (Table 1; Figure 1).Their breakpoints are shown in Figure 3.

Discussion
Most of the information obtained in this study is in accordance with the literature (review in Wasserman, 1992).Fixed inversions on chromosome 4 were found in D. onca as well as in D. carolinae.
What does not match previous findings is the fact that the 3c inversion in fasciola does not have the same 3c breakpoints as described in the mulleri complex by Wasserman (1962a, b).Accordingly, we propose that the denomination of the inversion present in the species of the mulleri complex be changed from 3c to 3u.This observation changes the previously proposed phylogenetic relationships among the species in the mulleri and fasciola subgroups as well as the relationships of the species in the fasciola subgroup within the repleta group.The fasciola subgroup becomes derivative of Primitive I in the repleta group and is no longer a derivative of the mulleri complex, as proposed by Wasserman (63) (Figure1).The present hy-  pothesis was required for the subgroup fasciola to be considered ancestral of the repleta group with a forest origin, as proposed by Throckmorton (1975Throckmorton ( , 1982) ) and "supported" by morphological and data (Pipkin, 1965;Pipkin et al., 1966;Sene et al., 1977;Pereira et al., 1983;Morais et al., 1995).This new phylogenetic hypothesis, based on chromosomal inversions, offers a better perspective to understanding the relationships within the subgroup to be inferred from other markers as in Costa and Sene (2002).Furthermore, based on cytological data, we propose two new species complexes within the fasciola subgroup: the fascioloides complex, comprising D. ellisoni and D. fasciloides which share one inversion (2d 3 ), two centric fusions (2-4F and 3-5F), and a great karyotype similarity regarding the X chromosomes (Kuhn et al., 1995); and the onca complex, comprising D. carolinae and D. onca, which share one inversion (4f 2 ) in addition to the similarity in the morphology of their aedeagi as shown by Vilela (1983).The species D. rosinae directly derives from Primitive VII, the hypothetic ancestral sequence to the fasciola subgroup (Figure1), after fixation of the 3c inversion.

Figure 1 -
Figure1-The diagram illustrates the revised phylogenetic relationships among some species within the fasciola subgroup and between the fasciola subgroup and the mulleri subgroup in the repleta group.as proposed in this study (modified fromWasserman, 1992:522).

Table 1 -
List of strains, collection localities and chromosomal arrangements of seven species of Drosophila in the fasciola subgroup.