Large wood recruitment in the Tegnas torrent (Italy): The impact of the Vaia storm and the role of the riparian forest structure

This study investigates large wood (LW) recruitment from the floodplain to the channel in a mountain stream in northeastern Italy, following the exceptional 2018 flood triggered by the Vaia Storm, a severe windstorm and intense precipitation event. It aims to quantify in-channel LW loads, estimate floodplain-recruited LW volumes, explore the links between hydraulic forcing, sediment balance, lateral connectivity and recruitment and explore the potential of numerical modelling in such a context. The study focuses on a 9.5 km segment of the Tegnas Torrent, a mountain stream with a catchment area of 52 km². Post-event in-channel LW was quantified through field surveys across sampling segments, while to assess recruitment volumes, a combination of pre- and post-event remote sensing data and 2019 field plots was used to estimate standing volumes and identify trees eroded by flood-induced channel widening. At the reach scale, key variables related to hydraulic forcing, sediment dynamics and lateral connectivity were evaluated, and correlations with LW recruitment were analysed. Finally, a two-dimensional numerical model was applied to simulate and compare the flood-driven erosion and wood recruitment. The total in-channel LW volume was estimated at 496 m³ ± 220 m³, averaging 18 m³ ha⁻¹, which is consistent with values from nearby, although undisturbed, mountain streams. In contrast, 2080 m³ (132 m³ ha⁻¹) of wood was recruited due to lateral channel widening. Our findings revealed that recruitment is influenced by complex factors, with limited correlation to sediment dynamics and hydraulic energy. Wider lateral erosion does not always lead to higher recruitment, as the riparian corridor's forest composition and structure play a key role. The numerical model provided reasonable estimates of channel widening and associated LW recruitment, making it a useful tool for approximating potential flood-induced planform changes. However, for more accurate results, further refinement in vegetation and sediment transport modelling is necessary.