Theory proposes that genomic admixture between formerly reproductively isolated populations can generate phenotypic novelty for selection to act upon. Secondary contact may therefore be a significant promoter of phenotypic novelty that allows species to overcome environmental challenges and adapt to novel environments, including during adaptive radiation. To date, this has largely been considered from the perspective of interspecific hybridization at contact zones. However, it is also possible that this process occurs more commonly between natural populations of a single species, and thus its importance in adaptive evolution may have been underestimated. In this study, we tested the consequences of genomic introgression during apparent secondary contact between phenotypically similar lineages of the riverine cichlid fish Astatotilapia calliptera. We provide population genetic evidence of a secondary contact zone in the wild, and then demonstrate using mate-choice experiments that both lineages can reproduce together successfully in laboratory conditions. Finally, we show that genomically admixed individuals display extreme phenotypes not observed in the parental lineages. Collectively, the evidence shows that secondary contact can drive the evolution of phenotypic novelty, suggesting that pulses of secondary contact may repeatedly seed genetic novelty, which when coupled with ecological opportunity could promote rapid adaptive evolution in natural circumstances.

In the original publication of the article, the author group was published incorrectly. The correct author group is given in this erratum.

General rightsThis document is made available in accordance with publisher policies. Please cite only the published version using the reference above. Full terms of use are available: http://www.bristol.ac.uk/pure/about/ebr-terms 3.The results revealed striking differences among the three species in spatial genetic structuring. The 31 river-lake migratory mpasa showed only weak yet significant population genetic structure within the 32 main Lake Malawi catchment, suggesting that there is no strong natal homing. The habitat generalist 33 sanjika showed only weak spatial genetic differentiation at microsatellite loci within the Lake Malawi 34 catchment, but moderate structure in mitochondrial DNA, potentially reflecting male biased dispersal. 35The river restricted dwarf sanjika showed strong genetic structure in both microsatellite and 36 mitochondrial DNA, suggesting strictly limited dispersal at both adult and juvenile stages. "Endangered" species endemic to the lake catchment, implying that conservation initiatives operating 43 at both local and catchment scales are needed to reverse local population decline.

General rightsThis document is made available in accordance with publisher policies. Please cite only the published version using the reference above. Full terms of use are available: http://www.bristol.ac.uk/pure/about/ebr-terms It has been proposed that the fish faunas of African rivers assemble through multiple 20 colonisation events, while lake faunas form additionally through intralacustine speciation. 21While this pattern has been established for many lineages, most notably cichlids, there are 22 opportunities to further investigate the concept using phylogenies of congeneric endemic 23 species within ancient lake catchments. The Lake Malawi catchment contains three river-24 spawning cyprinids of the genus Opsaridium, two of which are endemic. These species differ 25 in body size, migratory behaviour and habitat use, but it has never previously been tested if 26 these represent a monophyletic radiation, or have instead colonised the lake independently. We 27 placed these species in a broader phylogeny of Opsaridium and the related genus Raiamas, 28 including all known species from the river systems surrounding Lake Malawi. Our results 29suggest that each of the species has independently colonised the lake catchment, with all three 30 taxa having well-defined sister taxa outside of the lake, and all sharing a common ancestor 31 ~14.9 million years ago, before the Lake Malawi basin started to form ~8.6 million years ago.