The English name Yellow Wagtail is commonly used to describe the polytipic species Motacilla flava. More specifically it is also used for the British breeding race M.f.flavissima, and various English names are applied to other races: e.g. Blue-headed Wagtail M.f.flava, Ashy-headed Wagtail M.f.cinereocapilla, Grey-headed Wagtail M.f.thunbergi, Black-headed Wagtail M.f.feldegg, Sykes's Wagtail M.f.beema. As the former name is ambiguous and the latter ones would be unnecessarily ponderous when describing a group of birds of several races. I have tended to use Latin epithets throughout this thesis. However, this does not allow an effective plural noun to be applied, so in addition I have used the term "Wagtails", which, unless otherwise qualified, always refers to the species Motacilla flava. This also allows a degree of variety to be introduced to the text. African savanna regions have a climate which is markedly seasonal, although these seasons do not directly correspond with those experienced in temperate regions. Nevertheless, in the regions in which this work was undertaken, the dry season occurs from about November to March and is a period of relatively severe conditions for many bird species. Consequently I bave tended to refer to this period as winter, which should be taken as synonymous with dry season. As very few studies of the overall biology of Palaearctic migrants wintering in Africa have been previously unuertaken, I have attempted to make this account as comprehensive as possible. Inevitably this makes the text rather long, and I have therefore presented as much data as possible in the form of diagrams, which allow a greater amount of information to be presented in a small space.

M.Sc. / During the last decade, Cape Wagtails (Motacilla capensis) have been observed with several forms of feet abnormalities, such as missing toes and clubbed feet. The aim of this research was to find the reason for these abnormalities. The initial hypothesis was that ‘Feet abnormalities in the Cape Wagtail (M. capensis) are caused by the internal action of contaminants (e.g. pesticides and metals) through direct contact and/or secondary poisoning’. Wagtails were caught with mist nets and/or ground traps depending on the area and accessibility of the sampling sites. Wagtails with feet abnormalities, as well as those without, were caught at six localities around South Africa. Indicators of pollutant exposure and effect were studied. The bioaccumulation of certain metals (measures of exposure) in the feathers such as aluminium (Al), arsenic (As), cadmium (Cd), chromium (Cr), cobalt (Co), copper (Cu), iron (Fe), Pb, nickel (Ni), manganese (Mn), silver (Ag), strontium (Sr) and zinc (Zn) and Al, As, calcium (Ca), Cd, Cr, Co, Cu, Fe, Pb, Ni, Mn, Ag and Zn in the blood plasma were analyzed on the Inductively Coupled Plasma Mass Spectrometry (ICP-MS). Biomarker testing (measures of effect) was carried out in the blood such as acetylcholinesterase (AChE) which tested for nerve transmission inhibition, ä-aminolevulinic acid dehydratase (ALA-D) which tested for lead (Pb) bioaccumulation, catalase (CAT) which tested for the invasion of oxyradicles and oxidative stress and DNA damage which tested for DNA band length alteration caused by stress due to xenobiotics. Biomarker outcomes and metal analysis results were carried out on a site comparison basis. Acetylcholinesterase, ALA-D and DNA damage were found not to be linked to the occurrence of feet deformities at these sites. However, CAT revealed stress at all the sites which may have indicated the presence of a stress (or stresses) which could have been caused by feet abnormalities or other factors. A non-metric multidimensional scaling graph (NMDS) ordination suggested that biomarkers were not grouped according to sites that had affected birds and those that did not, but were rather assorted. Consequently, few real differences in biomarker responses were noted between healthy and affected birds. Catalase enzyme activity and DNA damage were suggested to be indicative of feet abnormalities although results were not convincing. Secunda displayed the highest concentrations of metals in the feathers when compared to the other sites. Most metals were higher than reference values for contaminated areas and recent South Africanreported concentrations. Almost all of the metals tested at each site exceeded the Ecological Quality iii Objectives (EcoQOs) for metals in bird feathers, based on Weavers (Ploceus sp.) in the Gauteng region. Zinc levels at all the sites posed some concern. Chromium and Ni were suspected to possibly trigger feet and toe abnormalities in Cape Wagtails. Levels of various metals in the plasma were higher at sites that had no deformities. This suggested that these metal concentrations did not cause or influence feet or toe deformities. Calcium levels were found to be lower at sites with affected birds which indicated that the deformities may perhaps be caused by decreased Ca concentrations. Metal concentrations were higher than the uncontaminated reference levels, but were lower than contaminated site reference values. / Prof. V. Wepener

<p>This thesis relates modern speciation theory to the effects of sudden changes in the range and size of populations. Special reference is made to the climatic oscillations during the Quaternary ice ages. A meta-analysis of laboratory experiments showed that support for allopatric speciation is weak, especially for the peripatric type of allopatric speciation. Furthermore, factors traditionally believed to increase the likelihood of speciation have had little effect on the generation of reproductive isolation in speciation experiments. However, the method of testing reproductive isolation appeared important, in the sense that experimentally derived sister populations were likely to demonstrate reproductive isolation from each other but not from the unaffected mother population. Raw data from mating tests showed that the poor isolation between mother and daughter populations was an effect of asymmetric mate preferences towards males from the mother population. This suggests that peripatric speciation can be effective in generating reproductive isolation between sister populations. The proposed mechanism is that males become less attractive to females by losing certain secondary sexual traits during population bottlenecks, and that females shift their preferences towards other male traits. Support for this mode of speciation is found in the widespread bird genus <i>Motacilla</i> (wagtails). This genus is characterised by extensive plumage variation and contains a large number of widely distributed taxa in the northern parts of its distribution. This thesis shows that taxonomic diversity of wagtails is inversely related to complexity in song and to diversity in molecular and mitochondrial markers. The northern taxa seem to be descendants of southern populations, which were subjected to bottlenecks during expansions into re-opened habitats after the last ice age. The bottlenecks would not only reduce genetic diversity but also inhibit cultural transmission of song to the leading edge of colonisers, allowing sexual selection on other traits, such as plumage. Rapid plumage differentiation among wagtail taxa appears to be a recurrent process and has lead to convergent evolution, making the currently recognised species <i>Motacilla flava</i> (Yellow Wagtail) polyphyletic.</p>

This thesis relates modern speciation theory to the effects of sudden changes in the range and size of populations. Special reference is made to the climatic oscillations during the Quaternary ice ages. A meta-analysis of laboratory experiments showed that support for allopatric speciation is weak, especially for the peripatric type of allopatric speciation. Furthermore, factors traditionally believed to increase the likelihood of speciation have had little effect on the generation of reproductive isolation in speciation experiments. However, the method of testing reproductive isolation appeared important, in the sense that experimentally derived sister populations were likely to demonstrate reproductive isolation from each other but not from the unaffected mother population. Raw data from mating tests showed that the poor isolation between mother and daughter populations was an effect of asymmetric mate preferences towards males from the mother population. This suggests that peripatric speciation can be effective in generating reproductive isolation between sister populations. The proposed mechanism is that males become less attractive to females by losing certain secondary sexual traits during population bottlenecks, and that females shift their preferences towards other male traits. Support for this mode of speciation is found in the widespread bird genus Motacilla (wagtails). This genus is characterised by extensive plumage variation and contains a large number of widely distributed taxa in the northern parts of its distribution. This thesis shows that taxonomic diversity of wagtails is inversely related to complexity in song and to diversity in molecular and mitochondrial markers. The northern taxa seem to be descendants of southern populations, which were subjected to bottlenecks during expansions into re-opened habitats after the last ice age. The bottlenecks would not only reduce genetic diversity but also inhibit cultural transmission of song to the leading edge of colonisers, allowing sexual selection on other traits, such as plumage. Rapid plumage differentiation among wagtail taxa appears to be a recurrent process and has lead to convergent evolution, making the currently recognised species Motacilla flava (Yellow Wagtail) polyphyletic.

<p>I use morphological, vocal, molecular, behavioural, ecological and distributional data to re-evaluate the systematics of three passerine bird groups, the <i>Mirafraassamica </i>complex (bush-larks), the genus <i>Seicercus</i> ("spectacled-warblers"; with emphasis on the the <i>S. burkii</i> complex) and the genus <i>Motacilla</i> (wagtails). Two new species are described: <i>Seicercus soror</i> and <i>Motacilla samveasnae</i>. I propose that the polytypic species <i>M. assamica</i> should be treated as four separate species: <i>M. assamica</i>, <i>M. affinis</i>, <i>M. microptera</i> and <i>M. marionae</i> (it is also remarked that the proper name of the latter is <i>M. erythrocephala</i>). That is primarily supported by vocalisations and mitochondrial DNA. The latter data set also suggests that <i>M. assamica</i> sensu lato is paraphyletic, since <i>M. erythroptera</i>, which is always treated as a separate species, is nested within the <i>M. assamica</i> complex. I propose that the polytypic species <i>S. burkii</i> comprises six sibling species. Some of these are found to breed sympatrically, although mainly or entirely segregated altitudinally. Mitochondrial DNA suggests that the <i>S. burkii</i> complex is non-monophyletic, and also that the divergence of the different taxa is much older than indicated by morphological and vocal data. According to the molecular phylogeny, both the genera <i>Seicercus</i> and its assumed sister genus <i>Phylloscopus</i> are paraphyletic. That is corroborated by independent data. The phylogenetic study of the genus <i>Motacilla</i> reveals incongruence between mitochondrial DNA, nuclear DNA and non-molecular data. I conclude that the nuclear gene tree reflects the organismal phylogeny more faithfully than the mitochondrial gene tree. The latter is likely to have been affected by introgressive hybridisation, possibly also stochastic lineage sorting. The most remarkable result that is strongly supported by both mitochondrial and nuclear DNA is that <i>M. flava</i> is non-monophyletic.</p>

I use morphological, vocal, molecular, behavioural, ecological and distributional data to re-evaluate the systematics of three passerine bird groups, the Mirafraassamica complex (bush-larks), the genus Seicercus ("spectacled-warblers"; with emphasis on the the S. burkii complex) and the genus Motacilla (wagtails). Two new species are described: Seicercus soror and Motacilla samveasnae. I propose that the polytypic species M. assamica should be treated as four separate species: M. assamica, M. affinis, M. microptera and M. marionae (it is also remarked that the proper name of the latter is M. erythrocephala). That is primarily supported by vocalisations and mitochondrial DNA. The latter data set also suggests that M. assamica sensu lato is paraphyletic, since M. erythroptera, which is always treated as a separate species, is nested within the M. assamica complex. I propose that the polytypic species S. burkii comprises six sibling species. Some of these are found to breed sympatrically, although mainly or entirely segregated altitudinally. Mitochondrial DNA suggests that the S. burkii complex is non-monophyletic, and also that the divergence of the different taxa is much older than indicated by morphological and vocal data. According to the molecular phylogeny, both the genera Seicercus and its assumed sister genus Phylloscopus are paraphyletic. That is corroborated by independent data. The phylogenetic study of the genus Motacilla reveals incongruence between mitochondrial DNA, nuclear DNA and non-molecular data. I conclude that the nuclear gene tree reflects the organismal phylogeny more faithfully than the mitochondrial gene tree. The latter is likely to have been affected by introgressive hybridisation, possibly also stochastic lineage sorting. The most remarkable result that is strongly supported by both mitochondrial and nuclear DNA is that M. flava is non-monophyletic.