自然选择进化过程可能很复杂，但又可以预测

1. 信息

2. 摘要

Despite evolution by natural selection being a deterministic process well-established across both science and society, the extent to which evolution can be predicted remains debated. A large body of previous work has usefully highlighted how individual selective pressures and stochastic events can each affect the predictability of evolution in isolation. Here, we argue that predicting evolution can be aided by understanding interactions between different sources of selection, termed ‘complex selection’, rather than simply considering their individual or summed effects. For example, selection can be affected by external perturbations such as seasonal or rare climatic events, by factors internal to a system such as the frequency of traits in a population or population density, and also by interactions between these and other processes. We outline how understanding complex selection can be central to predicting directional evolution, stasis and a range of dynamics, including the coupling of ecological and evolutionary dynamics that shape biodiversity. The framework outlined here is derived from insights across a range of study systems and argues that the prediction of evolution can be feasibly improved through nuanced understanding of complex selection. Evolution by natural selection is a well-evidenced and well-understood process, but predicting the course that it will take in the future is less clear. In this Perspective, Nosil and Gompert argue that considering the complex interactive effects of selection pressures can strengthen predictive frameworks and inform on evolutionary dynamics.

3. 图表

4. 引用

@article{nosil2026,
  author = {Patrik Nosil and Zachariah Gompert},
  title = {Evolution by natural selection can be complex yet predictable},
  journal = {Nature Reviews Biodiversity},
  year = {2026},
  doi = {10.1038/s44358-026-00200-x},
  publisher = {Springer Science and Business Media LLC},
}