Evaluating the influence of crop seasonality on flood-regulating ecosystem services in small river basins

Flood-Regulating Ecosystem Services (FRES) are widely used to assess the capacity of the environment to retain water during a storm, mitigate runoff and ultimately reduce flood risks within a river basin. FRES are commonly evaluated by assuming a change in land use to compare differences in the runoff of two scenarios: a baseline, representing actual conditions, and a scenario in which the land use of the environment is changed to barren ground. Agricultural areas contribute to flood regulation as they have a lower runoff with respect to barren or urbanised landscapes. However, methodologies to evaluate FRES in agricultural areas usually do not consider the variations in soil moisture that result from crop rotation throughout the year. Moreover, land use data typically used in such assessments describe with little or no detail the type of crop present in a given area.

To overcome these limitations, we introduce a methodology to evaluate seasonal FRES with two main research questions: i) is there a seasonality in the FRES of small agricultural river basins? ii) can different soil moisture conditions due to different crops have a diverse flood-regulating potential at the river basin scale?

The proposed approach is based on coupling two hydrological models: Watneeds, an agro-hydrological model that estimates daily soil moisture based on agricultural water demand, and Mobidic, a fully distributed rainfall-runoff model. Mobidic uses the soil moisture conditions derived from Watneeds as the initial state to simulate flood hazards during extreme storm events.

The methodology is applied in the upper Ombrone Grossetano river basin (Tuscany, Central Italy), where agricultural land constitutes a great part of the river basin area. The study used gridded datasets and ground observations for model calibration and analysis. Specifically, the Chirps dataset was bias corrected using ground observations and supported hydrological balance calculations in Watneeds. In contrast, rain gauge data from the Regional Hydrological Service were used to perform frequency analyses of extreme rainfall events and derive the rainfall quantiles modelled in Mobidic.

The results reveal that different crops produce distinct soil moisture conditions under identical weather patterns, influencing flood hazards in varying ways. FRES show a seasonality, with the maximum value at the end of the growing season, especially for the tributaries of the river with an area generally less than 60 km². The FRES supplied by the agroecosystem each month is compared with the FRES demand, i.e. the monthly average peak discharge. The results demonstrate that agricultural practices and crop scenarios can result in diverse flood responses depending on the season, offering valuable insights for flood risk management in small agricultural river basins. They suggest that policies governing crop selection, irrigation schedules, and crop calendars should also consider their potential impacts on flood regulation.

This research is part of the FLORAES project, funded by the Premio Florisa Melone 2023, an initiative by the Italian Hydrological Society to foster independent research and collaboration among young Italian hydrologists.