Summary: Root exudation is a key plant function with a large influence on soil organic matter dynamics and plant–soil feedbacks in forest ecosystems. Yet despite its importance, the main ecological drivers of root exudation in mature forest trees remain to be identified. During two growing seasons, we analyzed the dependence of in situ collected root exudates on root morphology, soil chemistry and nutrient availability in six mature European beech (Fagus sylvatica L.) forests on a broad range of bedrock types. Root morphology was a major driver of root exudation across the nutrient availability gradient. A doubling of specific root length exponentially increased exudation rates of mature trees by c. 5‐fold. Root exudation was also closely negatively related to soil pH and nitrogen (N) availability. At acidic and N‐poor sites, where fungal biomass was reduced, exudation rates were c. 3‐fold higher than at N‐ and base‐richer sites and correlated negatively with the activity of enzymes degrading less bioavailable carbon (C) and N in the bulk soil. We conclude that root exudation increases on highly acidic, N‐poor soils, in which fungal activity is reduced and a greater portion of the assimilated plant C is shifted to the external ecosystem C cycle.

Soil organic carbon (SOC) turnover in subsoils is assumed to be limited to spatially restricted microsites where fresh substrate inputs occur. Vice versa, the growth and activity of microorganisms outside of such hotspots may be limited by easily available substrates. The apparent long-term stability of subsoil organic carbon could thus be a result of microbial inactivity in these vast “cold regions” outside of hotspots. The aim of this study was to obtain realistic data about the in situ distribution of microbial hotspots in deep soil using soil zymography for three extracellular enzymes on undisturbed soil slices sampled from 0 to 161 cm depth. The results showed that most enzyme-driven turnover processes were concentrated to small portions of <1 to 10% of the subsoil volume, while enzymes in the major part of subsoils were barely active. In a second step, soil slices were homogenously sprayed with ¹⁴C glucose, incubated for 2 weeks and again analyzed with soil zymography. After glucose application, enzyme activities greatly increased in non-hotspot areas, thus confirming that substrate availability limits microbial activity in most of the subsoil volume. This implies that substrate limitation is a controlling factor for SOC stability in subsoils, suggesting that SOC in non-hotspots is persisting over long time periods until substrate becomes available and increases microbial activity.