Incorporating landscape configuration metrics improves the accuracy of water-related ecosystem services models

Local and regional authorities in the EU prepare water management plans to ensure water resources are managed sustainability, taking into account a wide range of factors such as climate, landscape and human activities. These plans consider the role of water-related ecosystem services - the functions performed by natural areas that benefit humans through modifying water quality, flow and availability. For example, forests and wetlands can reduce run-off and enhance groundwater recharge. 

Water management plans typically consider the different land uses across an area in their predictive modelling, but may not take landscape configuration into account, for instance using information about the size, shape and connectivity of the patches of any particular land use. However, studies have shown that land use configuration has a significant effect on water-related ecosystem services. Researchers in this study note that there was limited guidance on how these effects should inform management and policy decisions, therefore they looked at how configuration data could be used in modelling – potentially informing land management for water-related ecosystem services. 

Focusing on an area of around 8,000 km² in the Arno river basin in Tuscany, Italy, they used data models to evaluate the importance of landscape configuration factors in the provision of ecosystem services. They considered nine landscape configuration metrics relating to the size, shape and distribution of land-use patches, and evaluated these against three indicators of water-related ecosystem service provision:  

• Water yield – the available water supply;  
• Run-off – water that runs over the ground (rather than soaking into it or running through watercourses); 
• Groundwater recharge – water flowing down into the groundwater supply. 

The researchers analysed the variation in each indicator and evaluated differences over time and between locations. They used data models alongside historical data from 2000 to 2020 to evaluate the importance of the landscape configuration factors. They did this by running the models once with the historical data, and then again with substitute values for one factor, generated at random. The difference in accuracy of the model’s prediction in each case represented how important that factor was. 

The researchers used two models:  

• The BIGBANG model, developed by the Italian Institute for Environmental Protection and Research, designed to make predictions at national and regional scales; 
• The SWAT (Soil and Water Assessment Tool) model, presented in the journal Ecohydrology in 2023, designed for high resolution analysis of complex watersheds. 

The results from both models indicated that landscape configuration factors were associated with prediction accuracy. In the SWAT model, significance ranged from an importance of 43% in accounting for temporal variation in water yield, to 14% in accounting for spatial variation in groundwater recharge. In the BIGBANG model it ranged from 13% importance for spatial variation in groundwater recharge to 2% for temporal variation in run-off.  

In modelling run-off, coniferous forest patches were especially significant, with an increase in the number of patches corresponding to a decrease in variability. In the case of groundwater recharge, the importance of broadleaved forests were noteworthy, which reduced variability in this indicator. For water yield, they highlighted that forest patches reduced variation less effectively when they were more fragmented across the landscape. 

The researchers presented detailed figures on the importance of each specific metric in each scenario (i.e. each combination of indicator, model and temporal/spatial variation). They highlighted the importance of certain metrics in certain scenarios. For instance, they identified the core area index (the proportion of total area that is inside a patch rather than on the edge) for broadleaf forests as a significant influence on spatial and temporal variation of run-off according to the SWAT model, and of spatial variation in water yield according to BIGBANG. 

The study showed ways in which the use of land use configuration metrics might influence management decisions. These included: 

• Prioritising afforestation activities that increase forest connectivity to stabilise run-off and reduce soil erosion; 
• Implementing agroforestry in areas with complex land-use patch shapes to improve groundwater recharge and reliability of groundwater supply; 
• Planning land use in ways that ensure a significant level of core area to improve water availability and yield. 

These results, said the researchers, demonstrated the potential value of incorporating land use configuration metrics into the models of water-related ecosystem services used to inform land-use planning and policy development. They argued for the development of models that would incorporate such metrics, and for their use in policy, planning and management.  

Some study limitations were noted, including the spatial resolution of the analysis. Further studies could consider the development of optimal landscape metrics and investigations of the role of specific metrics in predicting the behaviour of water-related ecosystem services. 

Reference: 

el Jeitany, J., Nussbaum, M., Pacetti, T., Schröder, B., Caporali, E. (2024) Landscape metrics as predictors of water-related ecosystem services: Insights from hydrological modelling and data-based approaches applied on the Arno River Basin, Italy. Science of The Total Environment 954: 176567 

http://doi.org/10.1016/j.scitotenv.2024.176567 

To cite this article/service:  

“Science for Environment Policy”: European Commission DG Environment News Alert Service, edited by SCU, The University of the West of England, Bristol.  

Notes on content:  

The contents and views included in Science for Environment Policy are based on independent, peer reviewed research and do not necessarily reflect the position of the European Commission. Please note that this article is a summary of only one study. Other studies may come to other conclusions.