Hybridization and adaptive radiations

As an iconic system in evolutionary biology, I’ve always been interested in African cichlids and the origins of their diversity1. These cichlids represent an adaptive radiation; they’ve evolved rapidly from a single origin to exploit and speciate into open niches (for a general overview of adaptive radiations, see Losos 2010). In the Great Rift Valley in East Africa, over 2,000 species of cichlids are present. In Lake Victoria region alone, there are more than 700 species that have evolved in only 150,000 years. The amount of diversity and the rate at which it has been generated is really quite incredible.

The cichlid adaptive radiations have received lots of attention, but it is still somewhat unclear how enough genetic variation could be present to generate the huge numbers of species so rapidly. More generally, how could one species possess the necessary variation to result in such diverse niche use?

There are two related, but distinct hypotheses concerning hybridization may explain the generation of diversity in African Cichlids. The first is that hybridization between two species may seed an adaptive radiation and provide the genetic variation necessary for the subsequent radiation events. That is, this initial hybridization event serves as the source of the entire radiation. Another idea is the “syngameon hypothesis”, which argues that hybridization between closely related lineages (i.e., incipient species in the adaptive radiation) can generate genotypes that allow previously unoccupied fitness peaks to be reached. The occasional hybridization between radiating lineages can then continue to facilitate the occupation of novel fitness peaks (Seehausen 2004 provides a thorough discussion of these topics).

Joana Meier and colleagues were interested in understanding the role that ancient hybridization may have played in the Lake Victoria adaptive radiation. In their recent manuscript, the authors obtained RAD-seq data from species within Lake Victoria and species throughout the rest of Africa. Assuming a single history across all loci, Lake Victoria species formed a distinct clade.

However, Patterson’s D statistics (Fig. 1) showed evidence for admixture from two closely related groups, Congolese and Upper Nile cichlids, into Lake Victoria cichlid species. Surprisingly, the amount of admixture from these two groups was equal in each Lake Victoria species. The directionality of this gene flow was inferred using a five population test, an extension of the D statistic. Finally, the authors used full genome resequencing to estimate ancestry block size. Blocks were small, around 3kb, suggesting that the admixture event was old, and there was high correlation in ancestry blocks between closely related species, though this correlation decreases with phylogenetic distance.

Figure 1: A. Genealogy for species used for D statistics. B. Results from D statistics. Values further from 0 indicate more gene flow. C. Schematic of gene flow causing positive (top tree) and negative (bottom tree) D statistics. Figure from Meier et al., 2017

These results support a hybrid swarm origin of Lake Victoria cichlids following hybridization between Congolese and Upper Nile cichlids. The equal admixture proportions from each parent population present in the Lake Victoria species supports one ancient hybridization event rather than introgression. The correlation in admixture tract also supports this conclusion. The authors also found that introgression was not greater between sympatric species, arguing against more recent introgression.

Figure 2: LWS opsin diversity due to hybridization. Orange points represent Lake Victoria species, red are Congolese, and blue are Nile species. Class I is associated with shallow clear habitats, class II is associated with deep turbid environments. Figure from Meier et al., 2017.

The authors then looked specifically at the long-wavelength sensitive (LWS) opsin gene. This gene plays a role in adaptation to light availability based on water depth and is probably associated with mate choice and sexual selection. There are two different haplotype clades associated with this genes; a deep and shallow clade. Interesting, the Congolese and Nile lineages are unique and basal to each lineage (Fig. 2). The authors state that,

“The finding that the two LWS opsin haplotype classes of the LVRS [Lake Victoria cichlid species] are each shared uniquely with one of the parental lineages suggests that the ancient Congo-Nilotic admixture event was the source of the high functional variation among LVRS cichlids at the LWS opsin gene”

I think that this study is pretty convincing in showing that an ancient hybridization event provided the necessary variation to seed the adaptive radiation we see today. Though, the question still remains as to how widespread this phenomenon is in other adaptive radiations.


  1. Cichlids also provide an excellent system over which to argue about sympatric speciation.



Losos, Jonathan B. “Adaptive radiation, ecological opportunity, and evolutionary determinism: American Society of Naturalists EO Wilson Award address.” The American Naturalist 175, no. 6 (2010): 623-639.

Meier, Joana I., David A. Marques, Salome Mwaiko, Catherine E. Wagner, Laurent Excoffier, and Ole Seehausen. “Ancient hybridization fuels rapid cichlid fish adaptive radiations.” Nature Communications 8 (2017).

Seehausen, Ole. “Hybridization and adaptive radiation.” Trends in ecology & evolution 19, no. 4 (2004): 198-207.

About Reid Brennan

I am an evolutionary ecologist and a PhD candidate at UC Davis. I am generally interested in mechanisms allowing populations to evolve and respond to environmental stressors. Specifically, I combine physiological and genomic approaches to understand how fish have evolved to inhabit divergent abiotic environments.
This entry was posted in adaptation, evolution, genomics, next generation sequencing, population genetics, speciation and tagged , , , . Bookmark the permalink.