Funded Grants

Adaptive speciation in spatially structured populations

Imagine a pristine rainforest where some time far back in the past you would encounter a species of birds of intermediate size and with a brownish plumage. If you came back 200,000 years later, the brown birds would have gone, and instead you would find two types of birds, a small one with reddish tail feathers and a large one with bluish tail feathers. If you were able to do a phylogenetic analysis, you would find that these two new groups are very closely related, and that the brown birds were their common ancestor. Naturally, you would wonder what happened to the brown birds during the 200,000 years that have gone by. The short answer is: speciation has occurred.

This, however, is only a name for a very complicated process that has occupied the minds of theoretical and empirical biologists ever since Darwin. While Darwin's main focus were the processes by which already existing species adapt to different niches (such as resources that are easiest to consume by small or large bids), he was well aware of the difficulties that arise in explaining the origin and persistence of species, e.g. with regard to the establishment of mating barriers and reproductive isolation (as when birds only mate with other birds having similarly colored tail feathers). In fact, for Darwin speciation was 'the mystery of mysteries'.

Not only is it difficult to explain single events of evolutionary diversification like the split of an ancestral species into two distinct descendant species, but the amount of diversity that has evolved over time is enormous. When a scientist collected beetles from a single tropical tree species, he found 682 different beetle species, 140 of which he estimated to be specialized to live exclusively on that single species of tree. Since there are ca. 50,000 different species of tropical trees, he concluded that there are ca. 7 million of specialist beetle species alone! In fact, the total number of species of animals and plants on earth has been estimated to lie between 10 and 100 million. Adding to that the fact that the number of extant species is estimated to constitute ca. 1% of the number of species that ever existed on the planet, one arrives at a truly amazing picture of the evolution of biological diversity.

There is a rather surprising logical connection between the enormous amount of extant diversity and a single event of evolutionary diversification: presumably, all extant species arose form a single common ancestor. Understanding the mechanisms and processes that lead to speciation is therefore of fundamental importance in biology.

For speciation to occur, two processes are required. Subpopulations of an ancestral species must diverge genetically, and these diverging subpopulations must become reproductively isolated. It is difficult to imagine how subpopulations can genetically diverge from each other if they occupy the same geographical range, because these subpopulations would continue to exchange genetic material. It is perhaps deceptively easy to envisage that populations evolving in separate locations would eventually become reproductively isolated due to genetic incompatibilities. Even though the potential genetic mechanisms leading to reproductive isolation between geographically separated populations are actually only poorly understood, these considerations have led to the dominant belief among evolutionary biologists that speciation occurs most often in allopatry, i.e. due to geographical isolation between subpopulations of a single ancestral species. Prominent evolutionists have forcefully supported this view and have in essence proclaimed that speciation under sympatric conditions, that is, in the absence of geographical isolation, does simply not happen.

However, views about speciation have started to change in recent years, both for empirical and theoretical reasons. On the one hand, molecular phylogenetic data have revealed many examples of closely related species that have a common ancestor and that almost certainly evolved in sympatry. For example, species flocks of cichlid fishes in small crater lakes in Cameroon evolved under sympatric conditions, yet often have a common ancestor. In addition, there have been a number of experimental studies showing how ecological interactions, such as competition for food, can create conditions for divergence between sympatrically occurring types. Finally, the famous example of adaptation to different host plants in apple maggot flies shows how reproductive isolation can evolve in sympatry if mating behavior is contingent on diverging ecological traits.

These findings have received strong theoretical support from a new theory of adaptive dynamics. This theoretical framework for studying evolution as a dynamical system in phenotype space was developed by Hans Metz and his colleagues at the University of Leiden. The phenomenon of evolutionary branching is one of the paradigmatic features of adaptive dynamics. During evolutionary branching, an evolving population first gradually evolves to a point in phenotype space at which selection becomes disruptive due to ecological interactions within the evolving population. Disruptive selection implies selection against intermediate types, and consequently the population splits into two diverging phenotypic clusters. Evolutionary branching has been shown to occur in models involving all basic types of ecological interactions: competition, predation, and mutualism. In addition, models for evolutionary branching have been extended to include mulitlocus genetics and the evolution of assortative mating and reproductive isolation, which has led to a the development of a general theoretical framework for studying ecological mechanisms of speciation under fully sympatric conditions.

Together with the empirical results documenting sympatric origins of many species, the theoretical plausibility of evolutionary branching presents a strong challenge to the classical theory of allopatric speciation. The general aim of the work I propose to do is to reconcile ecological mechanisms and geographical patterns of speciation by studying processes of evolutionary diversification in spatially structured populations.

I first propose to develop stochastic individual-based models for coevolving populations occupying spatially continuous regions in order to assess the role of spatial structure for adaptive divergence. Together with Ulf Dieckmann from the International Institute for Applied Systems Analysis I have started to construct individual-based models for a single, spatially structured population in which individuals compete for resources. These models have already revealed surprising results: due to local adaptation along an environmental gradient and to spatially localized competitive interactions, individuals are more likely to interact with other individuals of similar phenotypes. The models show that this effect increases the propensity for evolutionary branching. Moreover, when evolutionary branching occurs the two emerging species are spatially segregated. Thus, spatial structure not only facilitates diversification, but intrinsically symmetric processes of speciation (local frequency-dependent competition) can generate allopatric species distributions. This provides an entirely new perspective on the role of geographical structure for speciation processes. The proposed theoretical work consists of a full-scale investigation of evolutionary branching in spatially structured populations whose coevolution is driven by all basic types of ecological interactions. The main focus will be on investigating the general phenomena of facilitation of diversification and spatial pattern formation as a consequence of spatially localized interactions. The resulting models are complex stochastic systems of interacting individuals that are computationally very demanding. With the help of analytical approximations, these models will elucidate the importance of geographical structure for processes of speciation. Our theory will thus establish a long due synthesis of allopatric and sympatric speciation scenarios.

While empirical findings have shown that speciation under sympatric conditions is plausible, direct experimental tests of the ecological mechanisms of speciation are still largely lacking: the long evolutionary time scales involved make it difficult to observe the whole process of diversification form beginning to end. However, experimental evolution in microbial model systems offers a way out of the conundrum of long generation times. Indeed, rapid diversification has been observed in microbial microcosms. In collaboration with Michael Travisano from the University of Houston I have started to test the ecological mechanisms driving such diversification by using the bacterium Escherichia coli as experimental model system. We have obtained promising preliminary results showing that under sympatric conditions, multiple resources promote evolutionary diversification of a single ancestral strain into multiple strains specializing on different resources. This conforms to expectations from adaptive dynamics models of evolutionary branching, and I will extend these experimental studies to spatially structured bacterial populations. This will be achieved by varying environmental resource conditions gradually over a series of chemostat cultures. Local dispersal is imposed by exchanging fractions of bacterial populations between adjacent chemostats corresponding to similar environmental conditions. By varying system parameters such as the migration rate and the slope of the environmental gradient, it is then possible to assess the effect of spatial structure on the likelihood and the amount of diversification. It will be very interesting to test the conditions leading to spatial pattern formation of the diverging strains across the environmental gradient. In addition, determining the genetic changes underlying diversification will give important insights into the genetics underlying diversification.

In sum, I propose to study the adaptive mechanisms leading to evolutionary diversification in geographically structured populations. A combined theoretical and experimental approach will provide essential insights into the complex process that ultimately generates biodiversity: speciation, the mystery of mysteries.