Local and downstream cumulative effects of traditional meadow management on stream-water quality and multiple riparian taxa.

Water quality and riparian communities are among the most affected stream components by agriculture. However, little is known about the effects of riparian management for both aquatic and terrestrial taxa at different spatial scales. Here, we surveyed aquatic (diatoms) and terrestrial taxa (bryophytes, vascular plants, litter-dwelling snails, and ground and volant arthropods), to compare the abundance and richness of riparian taxa and chemical quality between reference and exposed sites in two stream reaches each of c. 3.5 km in northwestern Spain. Impacts in exposed sites were mainly due to traditional farming practices (TFPs), which included traditional meadow management, weirs built for now-unused water mills and sporadic timber harvesting. Therefore, we measured ten covariates and predictors related to the intensification of TFPs at local and within-stream scales and explored associations with riparian and water-quality measures to study the potential effects of TFPs in more detail. Reference and exposed sites did not differ significantly in water properties (diatom-biotic indices, conductivity, total organic carbon and nitrates), but exposed sites had less concentrations of soil metals Cd, Cu, Ni and Zn and less cover and richness of riparian trees, as inferred by the index QBR. Exposed sites had more volant insect decomposers and reference sites a greater abundance or richness of snails, ground predators and decomposers. Bryophyte richness was greater in reference sites. Our inferences may inform the joint cumulative downstream effects of weirs, meadow uses and riparian alterations but were generally consistent with most riparian taxa benefiting from having larger forested areas. Given the contrasting responses among taxa, we argue that land snails, terrestrial flies, and centipedes may be valuable additions to current riparian assessments mostly based on plants, beetles and spiders as indicator taxa. Our study also suggests caution when inferring farming impacts on streams from the surface area of pastoral land.