Search

Androgenesis is a rare form of asexual male reproduction found in disparate taxa across the Tree of Life. Phylogenetic analyses of mitochondrial genes suggest that androgenesis has arisen repeatedly in the Asian clam genus Corbicula. Two of these androgenetic species have been introduced to North America. Multiple lines of genetic evidence suggest that although nuclear recombination between these two species is rare, mitochondrial genome capture is a frequent consequence of androgenetic parasitism of heterospecific eggs. Egg parasitism may also rarely result in partial nuclear genome capture between closely related species of Corbicula, which provides a mechanism for the otherwise clonal species to avoid the deleterious effects of asexuality. Egg parasitism among congeners may explain why androgenesis has been maintained in Corbicula after fixation and has not yet led to population extinction. This mechanism also provides an explanation for the apparent multiple origins of androgenesis in Corbicula as seen on the mitochondrial DNA phylogeny. We suggest that a single androgenetic lineage may have repeatedly captured mitochondrial genomes (as well as portions of nuclear genomes) from various sexual species, resulting in several distinct androgenetic species with distantly related mtDNA genomes and divergent morphologies.

Restriction-enzyme analysis of mitochondrial DNA and protein electrophoresis were used to document patterns of gene flow across a hybrid zone between chromosomal races of Peromyscus leucopus. Chromosomal markers (three inversions) are such that individuals can be classified as potential F,'s, backcrosses, or parental types. Allozymic characterization of the hybrid zone is congruent with the chromosomal data (Stangl, 1986) and indicates an assymetrical distri- bution of markers, with the northeastern markers being distributed at a higher frequency into southwestern populations. Restriction patterns of mtDNA suggest that the two cytotypes may have had different evolutionary histories, and the distribution of haplotypes is concordant with other genetic markers used to identify the hybrid zone. Concordant changes in chromosomes, allozymes, and mtDNA suggest that the most viable hypothesis for the origin of the zone is secondary contact. A unique aspect of this study is that the same individuals were used for protein electrophoresis, mtDNA analysis, and chromosomal analysis. Thus, patterns of genetic variation can be interpreted free of any historical bias associated with samples collected at different times.

Within the Rhogeessa tumida-parvula complex, there are seven cytotypes involving diploid numbers of 30, 32, 34, 42, 44, and 52. Based on G-band analysis, the 30, 32, 34, and 44 forms differ from each other by centric fusions (13 different fusions were identified). Two 2n = 32 samples (one from Belize and one from Nicaragua) having essentially identical standard karyotypes differed from each other by eight fusion events. Cytotypes are allozymically distinct; however, other studies indicate that the cytotypes are morphologically very similar if not indistinguishable (LaVal, 1973; Baker, 1984). G-band chromosomal data are best interpreted as indicating that, within this complex, several species exist which should be maintained by a post-mating isolating mechanism resulting from meiotic problems in F1 individuals. Mus musculus (Capanna et al., 1977; Capanna, 1982; White, 1978) and the Rhogeessa tumida-parvula complex are unusual among mammalian species thus far studied in their pattern of chromosomal variation and presumptive mode of speciation. Strong evidence exists suggesting that in both groups populations differing by multiple centric fusions are biological species. It appears that the population and reproductive biology of Mus and Rhogeessa are quite different and any model that attempts to explain the mode of speciation in these two diverse taxa, must be compatible with their different biological characteristics.

The possibility that the Peromyscus maniculatus complex of deer mice arose via the process described as "centrifugal speciation" (Brown, 1957) has been discussed by Bowers et al., 1973, Lawlor, 1974, Baker et al., 1975, and Greenbaum et al., 1978. This study was designed to yield additional data on the cytogenetics of three species of this complex in order to evaluate the "centrifugal speciation" model as a viable explanation of the evolutionary origin of the complex. Peromyscus melanotis and P. polionotus are believed to be peripheral isolates from the stock that evolved into the species P. maniculatus (Blair, 1950, Bowers et al., 1973). P. maniculatus is a widespread species occurring in a variety of habitats, whereas P. melanotis is a high montane species found in coniferous forest and P. polionotus is a deep sand species found in grassy fields at lower elevations near sea level (see Hooper, 1968, for general distribution of these three species). It is probable that P. melanotis and P. polionotus are more closely related to P. maniculatus than they are to each other. Based on gross chromosome morphology, the karyotypes of P. polionotus and P. melanotis are more like each other than either is like the karyotypes of P. maniculatus. We have examined the G-banding and C-banding pattern of P. polionotus and compared it to the banding patterns of P. maniculatus and P. melanotis (Greenbaum et al., 1978) to determine whether this similarity in karyotype is reflected in homologous banding patterns or is a product of convergent evolution. If the chromosomal data for the species complex are to fit the centrifugal speciation model then the karyotypes of P. polionotus and P. melanotis must be most like the primitive condition and the karyotype of P. maniculatus must be the most derived. Additionally, the centrifugal speciation model would predict more chromosomal polymorphism and geographic variation in the "central stock" (P. maniculatus) and less in the peripheral isolates (P. polionotus and P. melanotis). Our study is designed to determine the chromosomal homologies in these species and thereby test the hypothesis that this group is the result of centrifugal speciation.

We studied 75 individuals of the Plains pocket gopher, Geomys bursarius, from eastern New Mexico, where the subspecies major and knoxjonesi hybridize. Each individual was examined for chromosome number, ribosomal DNA, mitochondrial DNA, and three protein systems for which reference parental populations were fixed for alternative alleles. Twenty individuals were indistinguishable from parental major, 14 individuals were indistinguishable from parental knoxjonesi, and 41 individuals had genotypes composed of combinations of character states that distinguish the two parental types. The parental types appear to represent discrete genetic entities that have restricted introgression across a narrow hybrid zone (width approximately 3 km, using the 20/80 criterion). Parental types overlap in geographic distribution near the center of the zone, and changes in mitochondrial DNA and the five nuclear markers are concordant across the zone. It is probable that there is premating isolation between knoxjonesi males and major females. The frequencies of individuals with certain genotypic combinations within our sample imply differential reproductive success of certain genotypes. We propose that F1 's and highly heterozygous males are sterile and that hybrid females are less fertile than parental females. These postmating factors, along with premating isolation for one of the reciprocal crosses, probably account for the restriction of gene flow across the contact zone. The structure of the zone can be explained by the "dynamic equilibrium" model.

Recently there has been a surge of interest in karyotypic variation in the vertebrates, particularly in mammals (Benirschke, 1969; Ford, 1970; White, 1973), and in the relationships of chromosomal modification to ecological adaptation and speciation. Research on karyotypic diversity in rodents, including pocket gophers (Geomyidae) (Thaeler, 1968; Patton and Dingman, 1968; Patton, 1970, 1972; Patton et al., 1972; Davis et al., 1971), Gerbillus (Zahavi and Wahrman, 1957), Spalax (Wahrman et al., 1969), Sigmodon (Zimmerman, 1970), and Peromyscus (Hsu and Arrighi, 1968; Lee et al., 1972), has demonstrated that many morphologically defined species are in fact pairs or complexes of allopatric or parapatric semispecies and species isolated to varying degree in part at least by chromosomal differences. Because extensive chromosomal reorganization can result in genetic isolation between populations as a consequence of hybrid sterility (White, 1969; Todd, 1970), it follows that speciation theoretically can occur without genic modification. Conversely, the recent discovery of homosequential species of Drosophila suggests that "it is possible for speciation and evolution to be based entirely on mutational changes occurring at the submicroscopic level" (Carson et al., 1967; Carson et al., 1970). It is therefore of interest to compare patterns of chromosomal and genic divergence in complexes of closely related species. In a group of species with a wide range of variation in chromosome morphology and number, is there a correlated pattern of genic differentiation? To answer this question, we have studied four morphologically defined species of pocket gophers of the Geomys bursarius group (Hall and Kelson, 1959; Baker and Williams, 1973), in which previous work has demonstrated an impressive range of karyotypic diversity (Davis et al., 1971; Kim, 1972; Baker et al., 1973). Of the four species of the Geomys bursarius group (Fig. 1), G. bursarius is the most widely distributed, occurring in the Plains Region, where numerous subspecies are currently recognized (Hall and Kelson, 1959). Geomys personatus and G. arenarius have moderate-sized ranges, and each is divided into several subspecies. G. personatus is parapatric with G. bursarius in southern Texas (Kennerly, 1959), and G. arenarius has a disjunct range on the sandy areas along the Rio Grande and associated tributaries in New Mexico, Texas, and Chihuahua. The fourth species, G. tropicalis, is confined to an area of old sanddune habitat of approximately 300 square kilometers near Altamira, Tamaulipas (Baker and Williams, 1973). Geomys tropicalis apparently is separated geographically from its nearest neighbor, G. personatus, which extends southward from the main part of its range in Texas on the narrow barrier islands of coastal Tamaulipas. Present information indicates a gap of 250 kilometers (Selander et al., 1962; Alvarez, 1963). On the basis of morphological evidence and geographic considerations, Russell (1968) has proposed that arenarius is cladistically allied with bursarius and that geographic isolation was imposed in a postWisconsin period of increasing aridity.