Plots showing changes in the length of European eel in the restoration and control reaches of the River Glaven during the study period. Top plot: line plot showing mean length with 95% confidence intervals in restoration and control reaches during pre-restoration (1), post-embankment removal (2–3) and post-re-meandering (4–7). Emanating vertical dotted lines denote embankment removal and re-meandering. Bottom plot: marginal effects plot from the linear model showing no significant impact of interaction of Reach and Treatment on eel length, where a = pre-restoration, b = post-embankment removal, and c = post-re-meandering. The error bars represent 95% confidence intervals of the estimated effects.