Forests are the largest carbon sink in terrestrial ecosystems, and the impact of nitrogen (N) deposition on this carbon sink depends on the fate of external N inputs. However, the patterns and driving factors of N retention in different forest compartments remain elusive. In this study, we synthesized 408 observations from global forest ¹⁵N tracer experiments to reveal the variation and underlying mechanisms of ¹⁵N retention in plants and soils. The results showed that the average total ecosystem ¹⁵N retention in global forests was 63.04 ± 1.23%, with the soil pool being the main N sink (45.76 ± 1.29%). Plants absorbed 17.28 ± 0.83% of ¹⁵N, with more allocated to leaves (5.83 ± 0.63%) and roots (5.84 ± 0.44%). In subtropical and tropical forests, ¹⁵N was mainly absorbed by plants and mineral soils, while the organic soil layer in temperate forests retained more ¹⁵N. Additionally, forests retained more ¹⁵NH₄⁺ than ¹⁵NO₃⁻, primarily due to the stronger capacity of the organic soil layer to retain ¹⁵NH₄⁺. The mechanisms of ¹⁵N retention varied among ecosystem compartments, with total ecosystem ¹⁵N retention affected by N deposition. Plant ¹⁵N retention was influenced by vegetative and microbial nutrient demands, while soil ¹⁵N retention was regulated by climate factors and soil nutrient supply. Overall, this study emphasizes the importance of climate and nutrient supply and demand in regulating forest N retention and provides data to further explore the impacts of N deposition on forest carbon sequestration.

We searched the Web of Science Database for peer-reviewed papers prior to February 22, 2023, using “retention” and “N-15” as the keywords. The following criteria were applied to filter the peer-reviewed papers: (1) Selection of ¹⁵N tracer experiments in forest ecosystems conducted in the field, excluding laboratory incubation or greenhouse experiments; (2) Selection of the absolute value of ¹⁵N retention obtained from the ¹⁵N tracer experiment, excluding the relative value; (3) Selection of ¹⁵N tracer experiments including N addition treatments, excluding other treatments such as fire, phosphorus (P) addition, potassium addition, etc. Due to limited data on litter layers and understory vegetation components (i.e., shrubs, herbs, and grasses), the ¹⁵N retention of litter layers was combined into organic soil ¹⁵N retention. Within the entire forest ecosystem, the ¹⁵N retention of understory vegetation was not consider, focusing instead on the ¹⁵N allocation among different plant organs (i.e., leaves, branches, stems, roots). Ultimately, 408 observations were obtained from 56 peer-reviewed papers, totaling 62 sites and 92 site-years. The study sites were distributed across North America (25 sites), Europe (14 sites), Asia (14 sites), South America (3 sites), Oceania (4 sites), and Africa (2 sites), covering tropical forests (5 sites), subtropical forests (10 sites), temperate forests (42 sites), and boreal forests (5 sites). Raw data for ¹⁵N retention of different ecosystem compartments were obtained from tables, figures, results, or supplementary information in the peer-reviewed papers. When data were presented in figures, specific values were extracted using Getdata software 2.22 (GetData, Kogarah, NSW, AUS).