Thomas Blankers is assistant professor of evolutionary ecology at the University of Amsterdam. He is interested in divergent evolution and speciation, particularly in relation to biological interactions across taxonomic levels (within species, between species/kingdoms) . Visit his website for more information.
How predictable is evolution?
This question has historically been important in evolutionary biology. For example, quantitative genetic theory, specifically the Breeder’s equation and the Robertson-Price Identity, is concerned with short-term predictions of trait evolution. Gould’s famous thought experiment deliberates the repeatability of major evolutionary changes over geological time scales by asking whether similar species would evolve again if we would replay the tape of life (Fig. 1). What these two approaches have in common, is that they are attempting to resolve which aspects of evolutionary change are deterministic (predictable), and which aspects are inherently random (unpredictable).
Beyond the importance of a predictive research program for advancing evolutionary theory, evolutionary forecasting is undoubtedly important for society as a whole. Pressing societal challenges such as climate change, biodiversity loss, and resistance evolution in human or agricultural pests and pathogens are archetypal evolutionary problems (Wortel et al. 2022). These challenges underline the importance of understanding when and where we can make meaningful predictions about evolutionary dynamics. Excitingly, we live in a time where increasing availability of data and computational resources are helping us uncover the limits to evolutionary predictability, as illustrated in several recent reviews.
Lassig et al (2017) argue that an evolutionary process is predictable to the extent that selective forces reduce the possible paths of phenotypic evolution to a single dominant path. Although predicting the underlying genetics may be rather difficult because many genetic trajectories lead to similar functional outcomes, phenotypic evolution may be constrained by positive selection favoring specific outcomes (and negative selection disfavoring alternative outcomes). Such constraints would lead to predictable dynamics, as has been seen in several microbial and viral systems adapting to environmental challenges.
Blount et al. (2018) reflect on microbial and viral systems in which either identical populations or historically diverged populations are experimentally followed, evolving in identical environments. The most famous of these is the Escherichia coli Long Term Evolutionary Experiment by Richard Lenski’s lab. Blount et al (2018) argue that repeatability of experimental outcomes and convergence in natural systems depend on the strength and direction of natural (or artificial) selection and the extent to which evolutionary outcomes are influenced by the evolutionary history (known as historical contingency). When less evolutionary history divides two populations or species or when outcomes are less contingent on details of history and selection pressures are similar, we can expect evolution to be more likely to repeat itself.
Historical contingency is an example of data-limited predictability. Nosil et al. (2020) argue that data limits present the challenges for the future of evolutionary predictions. Limits resulting from stochasticity (“random limits”), such as allele frequency changes due to genetic drift, cannot be overcome. In contrast, limitation due to information about environmental variation, selection pressures, genetic architecture, and historical conditions (“data limits”) can sometimes be remedied. By collecting more data, such as time series of communities and genotypes, and measurements of effects and distributions of causal genetic changes, we can parameterize increasingly more complex models. However, some data limits may be more difficult to circumvent, as detailed knowledge about historical conditions is often not available and there are limits to the number of parameters we can estimate and model.
Following a cross-disciplinary workshop on evolutionary predictions organized by the Origins Center, Wortel et al. (2022) present an overview of the many types of evolutionary predictions from dynamics in experiments and biotechnology to eco-evolutionary interactions in complex communities. The main goal of this review is to illustrate diverse methods and data used to face similar challenges in different research fields concerned with evolutionary forecasting. The review also explores many genetic factors as well as the role of ecological interactions in shaping evolutionary predictability. An emergent insight in the field of evolutionary predictions is that more complex systems may also be evolutionary more stable and thus more predictable. There is an inherent predictability of complex dynamics that is conceptualized in theory of critical transitions (Scheffer 2020). This theoretical framework could be guiding in linking how we can predict evolutionary changes in complex communities over longer timescales from dynamics at shorter time scales.
Evolutionary forecasting serves both applied and fundamental evolutionary research on systems varying from clonal bacteria in mono-species cultures to complex communities. There are inevitable limits on our ability to project traits, species, and communities into the future, whether due to inaccessible data or randomness, but progress at every step brings our society closer to sustainable solutions and makes headway in our understanding of evolutionary theory.
References:
Lässig, M., Mustonen, V. & Walczak, A. Predicting evolution. Nat Ecol Evol 1, 0077 (2017). https://doi.org/10.1038/s41559-017-0077
Blount, Z.D., Lenski, R.E. & Losos, J.B. (2018) Contingency and determinism in evolution: replaying life’s tape. Science, 655. https://doi.org/10.1126/science.aam5979
Nosil, P., Flaxman, S.M., Feder, J.L. et al. Increasing our ability to predict contemporary evolution. Nat Commun 11, 5592 (2020). https://doi.org/10.1038/s41467-020-19437-x
Wortel, M. T., Agashe, D., Bailey, S. F., Bank, C., Bisschop, K., Blankers, T., Cairns, J., Colizzi, E. S., Cusseddu, D., Desai, M. M., van Dijk, B., Egas, M., Ellers, J., Groot, A. T., Heckel, D. G., Johnson, M. L., Kraaijeveld, K., Krug, J., Laan, L. … Pennings, P. S. (2022). Towards evolutionary predictions: Current promises and challenges. Evolutionary Applications, 00, 1– 19. https://doi.org/10.1111/eva.13513
Scheffer, M. (2020). Critical transitions in nature and society (Vol. 16). Princeton University Press.