Does plant uptake or low soil mineral-N production limit mineral-N losses to surface waters and groundwater from soils under grass in summer?

•Soil mineral-N has been measured over a year at two grassland sites in the UK.•Rates of mineral-N production have also been measured in vegetation-free soils.•In summer, though soils were warm and moist, rate of mineral-N production was low.•The effect is attributed to low litter inputs in summer when grass is growing well.•Low mineral-N production in summer must be limiting N losses to fresh waters.

Summer minima and autumn/winter maxima in nitrate concentrations in rivers are reputedly due to high plant uptake of nitrate from soils in summer. A novel alternative hypothesis is tested here for soils under grass. By summer, residual readily mineralizable plant litter from the previous autumn/winter is negligible and fresh litter input low. Consequently little mineral-N is produced in the soil. Water-soluble and KCl-extractable mineral N in fresh soils and soils incubated outdoors for 7 days have been monitored over 12 months for soil transects at two permanent grassland sites near York, UK, using 6 replicates throughout. Vegetation-free soil is shown to produce very limited mineral-N in summer, despite the warm, moist conditions. Litter accumulates in autumn/winter and initially its high C:N ratio favours N accumulation in the soil. It is also shown that mineral-N generated monthly in situ in soil substantially exceeds the monthly mineral-N inputs via wet deposition at the sites.