An empirically derived riparian forest preference index for lowland stream macroinvertebrates

• A riparian forest preference (RFP) index for macroinvertebrates was derived. • Forest within 15 m of the stream best predicted species presence. • 70 % of the species preferred forest, emphasizing the need for forest restoration. • The RFP index distinguished between varying degrees of tree cover. • The RFP index outperformed indices related to in-stream conditions. Over the past centuries, forest degradation and agricultural expansion reinforced stressors, like nutrient enrichment, hydromorphological degradation, and toxic pressure, which caused severe biodiversity losses. Stressor-specific macroinvertebrate indices are primarily linked to in-stream environmental conditions, and often focus on individual traits, while an empirically derived index that directly assesses the influence of forest cover and quality on macroinvertebrate assemblages, is currently lacking in the spectrum of available indices. Therefore, we derived a riparian forest preference (RFP) index, focusing on Ephemeroptera, Plecoptera, and Trichoptera (EPT), well-studied groups due to their sensitivity to a variety of stressors. We collected EPT presence-absence data from 1,993 sites in the Netherlands, spanning the period from 2010 to 2022, and linked the data to land use at spatial scales ranging from 15 to 500 m around 250 m stream segments. Boosted regression trees (BRTs) showed that riparian forest was the strongest predictor of species presence, especially at the smallest spatial scale of 15 m around the stream. Species-level RFP (RFP spec ) values were derived for 47 of the species, with 70 % of the species exhibiting positive RFP spec values, indicating a preference for forested riparian zones. Assemblage-level RFP asbl index values, calculated for 1,420 validation samples, showed a strong response to forest cover, underscoring the importance of restoring riparian zones close to the stream. Indices related to in-stream conditions responded less strong, as they are only indirectly linked to forest cover. It is concluded that the RFP index can be used well to guide and monitor lowland stream riparian forest restoration measures to restore biodiversity.