This work investigated two important evolutionary processes - speciation and adaptive variation - in two chromosomally-distinct allopatric Natal Midlands populations of the vlei rat Otomys irroratus. The two populations, at Kamberg and Karkloof, differ in the presence of a tandem fusion between chromosomes seven and 12 in the Kamberg karyotype. Speciation studies considered possible reproductive isolating mechanisms. In studies of adaptive variation, socio-ecological characteristics of both populations were investigated. Data on breeding and postnatal development provided evidence of post-zygotic barriers. Interpopulation pairs had reduced breeding success compared to intrapopulation pairs, and some hybrids died before weaning. Surviving hybrids had reduced growth rates, and almost all were sterile. In tests of pre-mating reproductive isolation, ethological barriers were emphasized. Individuals preferred same-population mates, suggesting the existence of mate recognition, which was achieved by means of population-specific courtship behaviour and communication, particularly olfactory, tactile and visual cues. No evidence of population-specific acoustic signals was found, although acoustic cues were associated with agonistic interaction, complementing other communicatory cues to promote increased aggression during interpopulation pairings. Laboratory studies of behaviour and morphology and field work (trapping and habitat assessment) provided information about socio-ecological parameters. The Kamberg habitat was harsher than the Karkloof one, as revealed by differences in seasonal and spatial availability of food and·cover. Cover was the key determinant of the level of sociability of both populations. Sparse, patchy cover selected for a partially communal social system in Kamberg o. irroratus: females were intrasexually tolerant and males were intrasexually highly aggressive. This, in conjunction with male-biased sexual dimorphism, implied that mating was polygynous. Abundant, uniform cover selected for a dispersed social system in Karkloof o. irroratus: females were intrasexually less tolerant than males. Ritualized aggression between males and a low degree of male-biased sexual dimorphism suggested that male may have overlapping home ranges in nature and that mating is promiscuous. Females possibly mated with dominant males, however. Contrasting social systems suggest that adaptation to local environmental circumstances has occurred in allopatry, and that Kamberg and Karkloof o. irroratus are undergoing adaptive speciation. Post-zygotic and pre-mating reproductive barriers appear to have evolved independently in both populations, and could potentially impede gene flow between the populations should they become sympatric. The presence of the tandem fusion in the Kamberg karyotype which, together with genetically-determined factors, may have caused hybrid sterility, suggests that this population is a chromosomally-determined incipient sibling species. / Thesis (Ph.D.)-University of Natal, Durban, 1993.

The evolutionary relationships of woodrats (Neotoma) were elulcidated through phylogenetic analyses of mitochondrial DNA restriction site and allozyme data. DNA samples from eleven nominal species from the genus Neotoma and two outgroup taxa, Ototylomys phyttotis and Xenomys nelsoni, were cleaved using a suite of 17 Type II restriction endonucleases. Mitochondrial DNA restriction profiles were visualized following electrophoresis of restriction digests via methods of Southern transfer and hybridization with 32P- and digoxigenin-labeled mtDNA probes. Restriction mapping resulted in the identification of 37 unique mtDNA haplotypes among the woodrat taxa examined. Proteins representing 24 presumptive structural gene loci were examined through starch gel electrophoresis. Binary-coded allozyme data and allozyme frequency data were analyzed using PAUP and FREQPARS, respectively. Phylogenetic analyses of the mtDNA restriction site data incorporated three different character type assumptions: unordered binary characters, Dollo characters, and differentially weighted unordered characters employing the STEPMATRIX option of PAUP. Proposed phylogenies for Neotoma are based on majority-rule consensus trees produced using bootstrap procedures. Phylogenetic analyses of the woodrat data sets revealed a distinct dichotomy among populations of white-throated woodrats (N. albigula) suggesting the presence of cryptic species within that taxon. MtDNA and allozyme data support the specific status of N. devia as distinct from N. lepida, and additionally reveal the presence of a third cryptic species referable to N. intermedia among the desert woodrats. Phylogenetic analyses of the genetic data also suggest subgeneric status for the desert woodrats, which is in agreement with evidence from morphology. The genetic data revealed a sister group relationship between N. stephensi and samples of N. mexicana, suggesting the placement of N. stephensi into the N. mexicana species-group. Neotoma fuscipes and N. cinerea formed a monophyletic lineage basal to the remaining members of the subgenus Neotoma which supports the assignment of N. fuscipes to the subgenus Teonoma with N. cinerea. Although stringent, Dollo parsimony methods produced the best supported phylogenies among the species of Neotoma. The STEPMATRIX approach was unable to resolve species relationships within species-groups but clearly delineated the higher taxonomic levels between species-groups and subgenera.

Proponents of the recognition concept of species hold that isolating mechanisms, including chromosome rearrangements, play no role in speciation while the more commonly accepted biological species concept proposes that isolation mechanisms are instrumental in the formation of new species. Moreover, some adherents of the biological concept of species, reject the hypothesis that chromosomal rearrangements can be instrumental in causing reproductive isolation and, hence, speciation. Evidence to the causative role played in speciation by chromosome changes can be obtained from cytogenetic investigations of sibling species, in parallel with analyses of gene products, DNA polymorphism and premating behaviour. This study reports the results of a cytogenetic investigation of 97 specimens of the vlei rat 0. irroratus, from 18 South African localities, and 11 samples of the Angoni vlei rat 0. angoniensis from two geographically distant populations. All 0. angoniensis individuals showed a constant karyotype with 56 acrocentric chromosomes but extensive variation was detected in 0. irroratus. Five cytotypes could be recognized within the latter. In the south-eastern parts of its South African range, 0. irroratus had a diploid number (2n) of 30 chromosomes in whicll all autosomes were acrocentric (cytotype A) while further east (cytotype A2), the diploid number was 30-32 with, again, acrocentric autosomes, A further acrocentric cytotype (AI) with 2n = 24-27 occupied the southern and south-eastern slopes of the Drakensberg range. A type with 2n = 28-30 (cytotype B), with eight pairs of biarmed autosomes, was found in the southern Cape region while in the Cape of Good Hope and in the north-eastern parts of South Africa, 0. irroratus had 2n = 28 with only four pairs of biarmed autosomes (cytotype C). Most of the numerical changes were due to variation in the number of copies of Bchromosomes which were small, biarmed and partly heterochromatic. C-banding analysis revealed that the short arms of bianned autosomes were totally heterochromatic. On the other hand, G-banding patterns of acrocentric autosomes were, with two exceptions (AI and A2 types), similar in all cytotypes while G~banding of the long arms VII of biarmed chromosomes matched the pattern of their homologues in acrocentric cytotypes. A potentially heterotic rearrangement was detected in the Al localities where a unique acrocentric autosome was identified as the product of a fusion between chromosomes 7 and 12. The geographic distribution of these groups of karyotypes correlated, by Discriminant Function Analysis, with bioclimatic regions of South Africa. The Al cytotype was shown to occupy the coldest and wettest region of the montane Drakensberg while the B type is found in the hot area of the eastern Cape with an unpredictable rainfall pattern: group C occupies regions of intermediate climate. Gene product analysis was carried out using the novel approach of subjecting liver homogenates to "Western blotting". This method was first assessed at supraspecilic level using specimens of various southern African rodents, and allowed the generation of phylogenies essentially similar to those produced by allozyme studies of the same taxa. At intraspecilic level, immunobloHing analysis did not reveal synapomorphies congruent with karyotype groups. This was interpreted, in conjunction with available allozyme data from the same populations, as evidence of low genetic differentiation between 0. irroratus cytotypes, A measure of genetic divergence was indicated in two populations from the Cape province and this was in agreement with existing data from allozyme electrophoresis and mitochondrial DNA polymorphism. The cytogenetic results were related to available data on breeding and premating behaviour concerning some of the O. irroratus populations investigated here. The presence of the 7/12 chromosome fusion in the Al cytotype correlated with a dramatic reproductive impairment of FI individuals originated from Al/A2 and Al/B cytotype crosses. Evidence of partial premating behavioural barriers has been reported by others, but information on premating behaviour between populations which are not chromosomally isolated is lacking. Therefore, it was not possible to establish if behavioural premating barriers preceded, or followed, the fixation of negatively heterotic chromosomal rearrangements. It was, nevertheless, suggested that the existence of such impaired mate recognition may be an example of reproductive character displacement which may have followed the fixation of the t(7: 12) typiVIII cal of the Al populations. In conclusion, the existence of chromosome changes in the AI, and possibly A2, populations accompanied by low genetic divergence and severely impaired hybrid reproductive success, are consistent with a hypothesis whereby chromosomal reproductive isolation causes speciation. Nonetheless, other speciation mechanisms mediated by genetic divergence and/or mate recognition failure, are possible in other populations where no chromosome changes of negatively heterotic potential were found. / Thesis (Ph.D.)-University of Natal, 1996.