Novelty by Furcation and Fusion: How tree-like is evolution?

Abstract

Novelty and innovation are fundamental yet relatively understudied concepts in evolution. We may study novelty phylogenetically, with a key question of whether evolution occurs by tree-like branching, or through exchange of distantly related parts in processes akin to horizontal transfer. Here, I argue that except at the lowest levels of biological organization, evolution is not usually tree-like. Perfectly vertical inheritance, an assumption of evolutionary trees, requires simultaneous co-duplication of all the parts of a duplicating or speciating (which I collectively call 'furcating') biological feature. However, simultaneous co-duplication of many parts usually requires variational processes that are rare. Therefore, instead of being perfectly tree-like, evolution often involves events that incorporate or fuse more distantly related parts into new units during evolution, which herein I call 'fusion'. Exon shuffling, horizontal gene transfer, introgression and co-option are such fusion processes at different levels of organization. In addition to co-duplication, units under phylogenetic study must individuate (gaining evolutionary independence) before they can diverge. A lack of individuation erases evolutionary history, and provides another challenge to tree-like evolution. In particular, biological units in the same organism that are the products of development always share the same genome, perhaps making full individuation difficult. The ubiquity of processes that fuse distantly related parts or oppose individuation has wide ranging implications for the study of macroevolution. For one, the central metaphor of a tree of life will often be violated, to the point where we may need a different metaphor, such as economic public goods, or a web of life. Secondly, we may need to expand current models. For example, even under the prevailing model of cell-type evolution, the sister-cell-type model, a lack of complete individuation and evolution by co-option will often be involved in forming new cell-types. Finally, these processes highlight a need for an expansive toolkit for studying evolutionary history. Multivariate methods are particularly critical to discover co-variation, the hallmark of an absence of complete individuation. In addition to studying phylogenetic trees, we may often need to analyze and visualize phylogenetic networks. Even though furcation - the splitting and individuation of biological features - does happen, fusion of distant events is just as critical for the evolution of novelties, and must formally be incorporated into the metaphors, models, and visualization of evolutionary history.