Investigating ecological and genetic constraints to speciation

Adaptive radiations gave rise to some of the most species rich groups of organisms on Earth. In many young and species rich adaptive radiations, extant species vary in the extent of genomic differentiation, potentially representing different stages in the process of speciation. The species of many radiations also co-occur with one or several closely related clades that did not radiate. Due to these two characteristics, young adaptive radiations provide ideal systems to investigate the speciation process and what might constrain it. In my thesis, I investigate the speciation processes in an adaptive radiation and its relatives that did not radiate. I address the question why some lineages speciate whereas others do not, even though they co-occur in a very similar or even the same habitat. In chapter 1 I study what might be the very onset of an adaptive radiation and provide evidence for differentiation along several, potentially orthogonal axes, within a haplochromine cichlid lineage only 50 years after it invaded a habitat suitable for cichlids to speciate. I show that differential selection pressure along an ecological axis might be of importance for the morphological and ecological differentiation among some of the morphs. In chapter 2 I demonstrate that the prevalence of disruptive ecological selection between conspecific morphs and sister species can explain the stage of speciation within two species rich genera of haplochromine cichlids in Lake Victoria. Additionally, we found that sister species pairs that differ strongly in male nuptial colouration, an indicator for divergent sexual selection, were genomically more distinct from one another than those with similar colouration, given the same prevalence of disruptive ecological selection. In chapter 3 I investigate causes of lack of speciation in an “ancient” endemic haplochromine cichlid species of Lake Victoria that co-occurs with many members of the Lake Victoria radiation occupying the exact same habitats that resulted in the formation of many species in the radiation. I show that this ancient lineage evolved morphological, ecological and genomic differentiation between para-allopatric populations in a similar range as observed between allopatric sister species in the radiation. Yet, in this ancient lineage these differences do not result in speciation because they are not linked to reproductive isolation upon secondary contact. In chapter 4 I investigate the genomic, phenotypic and ecological variation and differentiation in 19 sympatric morph and species pairs of East-African cichlids, some of which belong to the Lake Victoria radiation, whereas others do not. While ecological habitat contrast seems to be a good predictor for speciation in sympatric pairs of the radiating lineage, it is not in the other lineages. I find that the same habitat contrasts that are associated with disruptive ecological selection in members of the radiation do not generate disruptive selection and therefore do not initiate speciation in other lineages. I propose that the considerable genomic variation, that is a legacy of past hybridisation, is an essential ingredient to unusually rapid speciation and adaptive radiation in sympatry. Together the results reported in this thesis reveal the intricate interactions of disruptive ecological selection, disruptive sexual selection, genomic variation and ecological opportunity in rapid speciation and thus also their potential to facilitate or constrain the processes of speciation and adaptive radiation.