The temporal distribution of directional gradients under selection for an optimum
The temporal distribution of directional gradients under selection for an optimum
复制标题
DOI：
10.1111/evo.12532
复制
发表时间：
2014-12-01
期刊：
影响因子：
3.3
通讯作者：
Haller, Benjamin C.
中科院分区：
文献类型：
--
作者：
Chevin, Luis-Miguel;Haller, Benjamin C.
表型选择的时间变化通常归因于环境变化导致与适合度相关的性状的适应面运动，但这种联系很少得到实证证实。波动表型选择可以通过方向选择梯度随时间的方差和自相关来测量。然而，这些梯度的动态变化不仅取决于环境变化对适合度表面的改变，还取决于表型分布的进化。因此，尚不清楚选择梯度的可变性能在多大程度上告诉我们其波动的潜在驱动因素。为了研究这个问题，我们推导了在一个有限种群中，一个表型最优选择下方向梯度的时间分布，该表型最优要么恒定，要么以各种方式随机波动。结合基于种群和个体的模拟，我们的分析结果表明，尽管可以区分一些特征模式，但最优（包括恒定最优）的非常不同类型的变化可以产生相似的选择梯度的时间分布，这使得很难推断出明显波动选择背后的过程。分析表型分布的变化和选择梯度的变化，对于推断估计的波动选择的机制更有用。
Temporal variation in phenotypic selection is often attributed to environmental change causing movements of the adaptive surface relating traits to fitness, but this connection is rarely established empirically. Fluctuating phenotypic selection can be measured by the variance and autocorrelation of directional selection gradients through time. However, the dynamics of these gradients depend not only on environmental changes altering the fitness surface, but also on evolution of the phenotypic distribution. Therefore, it is unclear to what extent variability in selection gradients can inform us about the underlying drivers of their fluctuations. To investigate this question, we derive the temporal distribution of directional gradients under selection for a phenotypic optimum that is either constant or fluctuates randomly in various ways in a finite population. Our analytical results, combined with population- and individual-based simulations, show that although some characteristic patterns can be distinguished, very different types of change in the optimum (including a constant optimum) can generate similar temporal distributions of selection gradients, making it difficult to infer the processes underlying apparent fluctuating selection. Analyzing changes in phenotype distributions together with changes in selection gradients should prove more useful for inferring the mechanisms underlying estimated fluctuating selection.