ABSTRACT Agroforestry, defined as farming with trees, is growing in popularity for its potential to deliver multiple landscape benefits, including natural flood management (NFM). Most NFM research has been concentrated in uplands, whereas optimising tree placement on farms for NFM in lowland catchments has not been studied. Using spatially distributed TOPMODEL, we consider rainfall‐runoff changes under various agroforestry planting scenarios in a 1.75 km 2 lowland catchment in Yorkshire, UK, asking whether tree planting at densities practical for farming can significantly alter flood peaks, and which placement strategies generate the most significant effect on flood risk reduction for a fixed area of afforestation. Riparian woodland covering 2.5% of farmed area (0.6% of the catchment) attenuated flood peak by 2.6% and delayed its arrival by 6.6% (19 min). Shelterbelt or in‐field copses had a more modest influence on flood peak, and linear features such as in‐field tree rows or hedges had a negligible or opposing effect. Increasing tree coverage generated a diminishing return and woodland over the whole farmed area (25% of the catchment) produced only 1.6× the peak size reduction as 0.6% riparian cover in low flow scenarios, rising to 13 times the reduction for higher flows, despite requiring 40 times more land taken out of agricultural production. Our modelling therefore suggests that agroforestry sited on shallow gradients or in the riparian zone contributes to NFM without substantially impacting land available for productive agriculture. These findings demonstrate the potential capacity of agroforestry to deliver multiple landscape benefits.
Judson et al. (Fri,) studied this question.