Supplementary Material

Fig. S1. Restoration reach of the River Glaven at Hunworth (north Norfolk, UK): (a) in January 2009, prior to the renaturation project; (b) after removal of embankments in March 2009; and (c) in December 2010, after recreation of meanders in August.

Fig. S2. Diagnostic plots suggesting that the assumptions of normally distributed, homoscedastic and independent residuals of linear model testing for effect of Time period on eel density at the control reach were not severely violated.

Fig. S3. Diagnostic plots suggesting that the assumptions of normally distributed, homoscedastic and independent residuals of linear model testing for effect of Time period on eel length at the control reach were not severely violated.

Fig. S4. Diagnostic plots suggesting that the assumptions of normally distributed, homoscedastic and independent residuals of linear model testing for effect of Time period on eel weight at the control reach were not severely violated.

Fig. S5. Diagnostic plots suggesting that the assumptions of normally distributed, homoscedastic and independent residuals of linear model testing for effect of interaction of Reach and Treatment on eel length were not severely violated.

Fig. S6. Diagnostic plots suggesting that the assumptions of normally distributed, homoscedastic and independent residuals of linear model testing for effect of interaction of Reach and Treatment on eel weight were not severely violated.

© J.D. Champkin et al., Published by EDP Sciences 2024