Ancient arid corridor and niche divergence shaped the Cape Robin-chat’s split populations

The Cape Robin-chat, a familiar songbird across southern Africa’s drylands, has long puzzled biologists with its strangely split range. One set of resident populations occupies the aridlands of South Africa, Namibia, Mozambique and Zimbabwe, while another lives more than a thousand kilometres away in the montane forests of Malawi, Tanzania, Kenya, Uganda, Rwanda, Burundi and the Democratic Republic of Congo. A new study, published in Frontiers of Biogeography, uses genetics, climate models and dispersal simulations to show that these two populations are connected by a one-way gene flow highway that ran along an intermittent arid corridor from the Horn of Africa to southern Africa, and that niche divergence has since locked the southern birds out of the return journey north.

The research tests what biogeographers call the “species pump” hypothesis at the level of a single species. This hypothesis proposes that much of southern Africa’s extraordinary bird diversity arose when arid-adapted lineages from northeastern Africa repeatedly colonized the south during the Plio-Pleistocene climatic fluctuations of the last few million years. As the Afrotropical forests contracted and the East African Savanna biome expanded, dry-adapted species could move south along a corridor stretching from the Horn of Africa to southwestern Africa. Previous work on birds such as fiscal shrikes and other aridland lineages has shown deep genetic structure between northern and southern populations, consistent with this scenario, but the mechanism by which such structure is generated and maintained has remained unclear.

The Cape Robin-chat is an ideal test case because it defies the neat categorization of most African birds as either forest or aridland specialists. In the south it is a habitat generalist, living in forest edges, parks and gardens from sea level up to 3,000 metres on the South African escarpment, and it has even colonized the extremely arid Nama Karoo biome within the past few hundred years. In the north, by contrast, it is a montane forest resident restricted to elevations between 1,500 and 3,400 metres. Because temperature, latitude and elevation are tightly linked, the northern and southern populations, which span more than 35 degrees of latitude and over 3,000 metres of elevation, may each be adapted to substantially different climatic environments.

To untangle this history, the researchers genotyped 335 individuals from all four recognized subspecies at 14 highly polymorphic microsatellite loci, building on an earlier dataset. They used Bayesian clustering to identify population structure, assigned individuals to ancestry classes with a program called NewHybrids, and estimated gene flow with a coalescent approach implemented in migrate-n. They also built environmental niche models with MaxEnt, hindcast them onto 62 downscaled paleoclimate snapshots spanning the past 120,000 years using the HADCM3 global circulation model, and ran explicit dispersal simulations with the KISSmig package to map accessibility from either a northern or a southern origin.

The genetic results were striking. The best-fit clustering identified three populations: one encompassing all the northern samples, and two within southern Africa. Notably, the northern cluster did not match the recognized subspecies C. c. iolaema and C. c. kivuensis, suggesting that subspecific taxonomy in the north does not reflect the true genetic structure. The deepest divide in the data separated northern from southern populations, with a sharp genetic break near the Mozambique-Tanzania border. Admixture analysis found 247 individuals assigned as southern parentals, 27 as northern parentals, and 15 as F2 hybrids, but no F1s or backcrosses, a pattern consistent with a long period of isolation between the two lineages.

Gene flow estimates revealed a highly asymmetric pattern: movement from north to south was roughly ten times greater than movement from south to north, with migration rates of 0.199 versus 0.013. Graph-theoretic analyses of genetic connectivity showed that north-south gene flow occurs primarily from the Democratic Republic of Congo and Malawi into the South African grasslands populations, while western South African populations exchange genes only indirectly through the grasslands or Namibian populations. This directional pattern is exactly what the species pump hypothesis predicts, with northern lineages seeding southern populations but rarely the reverse.

The paleoclimate models added a temporal dimension. Hindcasting the niche models across 62 time slices revealed highly dynamic suitability, with dramatic contractions and expansions in both the northern and southern portions of the range, and intermittent connectivity between them. Static refugia were predicted in South Africa, along the Zimbabwe-Mozambique, Tanzania-Malawi and Democratic Republic of Congo-Tanzania borders, and in Kenya, but notably there were no stable areas predicted between the northern and southern regions. Dispersal simulations integrated over the past 120,000 years showed two connected areas, one in the south encompassing South Africa, southern Namibia, Zimbabwe and Mozambique, and another in the north encompassing the Democratic Republic of Congo, Malawi, Kenya and Tanzania, separated by a disjunction at the Mozambique-Tanzania border.

The niche analyses provided the crucial mechanistic insight. Environmental niche models built separately for the northern and southern populations revealed that the northern populations have a corridor of environmental suitability connecting them to the south, while the southern populations lack a comparable corridor northward. Niche identity tests were statistically significant for both Schoener’s D and Hellinger’s I, and niche overlap was low in both model-based and direct environmental space analyses, supporting niche divergence between the two populations. Interestingly, the background environments of the two regions were not significantly different, and the linear range break test found no evidence of a steep environmental gradient, meaning that the divergence is not simply a product of different geographic backgrounds but reflects genuine niche evolution.

The authors propose a modified species pump model in which niche divergence and subsequent dispersal limitation driven by environmental filters have contributed to population differentiation. Under this scenario, northern individuals dispersed south along the intermittent arid corridor, established the southern populations, and then, as the southern populations adapted to their new aridland environments, they lost the ability to move back north because the environmental corridor no longer suited them. This one-way filter, combined with periodic isolation during forest expansion, would generate and maintain genetic differentiation, and over evolutionary time could contribute to the rich avifaunal diversity of southern Africa.

The study also highlights the role of complex El Niño Southern Ocean oscillation dynamics in shaping dispersal routes throughout sub-Saharan Africa, a mechanism previously linked to mammalian movements that the authors expect would similarly affect many bird species. The genetic break identified here corresponds broadly with the East African suture zone recognized for mammals and with the division between two traditionally recognized montane regions of bird endemism, the Eastern Arc mountains and the Tanganyika Rift-southern Africa region. The authors suggest that, as with the Olive Sunbird, the Eastern Arc mountains may have served as a source for the southern African populations. They also note that biotic interactions with the congener Cossypha heuglini, which restricts C. caffra to montane habitat where the two species co-occur, may provide additional insights into the proximate causes of ecological divergence. Future work using genomic data to identify loci associated with environmental gradients, along with physiological assays and phenotypic comparisons, will be needed to fully resolve the eco-evolutionary dynamics at play, and to determine whether other bird species with similarly broad ranges share the same pattern of asymmetric connectivity and niche divergence across the Mozambique-Tanzania border.

Subject of Research: Population connectivity and niche divergence in the Cape Robin-chat across disjunct African ranges

Article Title: Niche dynamics modulate population connectivity between disjunct ranges of the Cape Robin-chat ( Cossypha caffra ) supporting an aridlands species pump

Article References: Wogan, G. O. U., Voelker, G., Jain, T., Kaliba, P., & Bowie, R. C. K. (2024). Niche dynamics modulate population connectivity between disjunct ranges of the Cape Robin-chat (Cossypha caffra) supporting an aridlands species pump. Frontiers of Biogeography, 17, Article e132679. https://doi.org/10.21425/fob.17.132679

Image Credits: AI Generated

DOI: 10.21425/fob.17.132679

Keywords: Cape Robin-chat, species pump, aridlands corridor, niche divergence, gene flow, paleoclimate, biogeography, southern Africa, population genetics, environmental niche models, Plio-Pleistocene, avian diversity