Statistical analysis
We used R (v4.2.1; R Core Team 2021) and RStudio (version 2023.3.0.386; Posit Team 2023) to carry out the statistical analysis. All data was checked for the presence of significant outliers and we used linear models to assess linearity, distribution of residuals and homogeneity of variance. Unless otherwise stated, no statistically significant outliers were detected and all data met assumptions of linearity, homogeneity of variance and normal distribution of residuals, and two-tailed correlations were tested with Pearson’s correlation coefficient. We used cluster-based permutation testing, using the “permutes” (Voeten, 2022) and “permuco” (Frossard, 2021) packages, to analyse instantaneous alpha. 5000 permutations of randomly shuffled data were used to create a distribution of cluster sizes expected under the null- hypothesis. Clusters of real data that exceeded the 95thpercentile of this null distribution were deemed statistically significant.