Does nitrogen transfer between plants confound 15N-based quantifications of N2 fixation?

Transfer of fixed N from legumes to non-legume reference plants may alter the N-15 signature of the reference plant as compared to the soil N available to the legume. This study investigates how N transfer influences the result of N-15-based N-2 fixation measurements. We labelled either legumes or non-legumes with N-15 and performed detailed analyses of N-15 enrichment in mixed plant communities in the field. The results were used in a conceptual model comparing how different N transfer scenarios influenced the N-15 signatures of legumes and reference plants, and how the resulting N-2 fixation estimate was influenced by using reference plants in pure stand or in mixture with the legume. Based on isotopic signatures, N transfer was detected in all directions: from legume to legume, from legume to non-legume, from non-legume to legume, from non-legume to non-legume. In the scenario of multidirectional N transfer, N-2 fixation was overestimated by using a reference plant in pure stand. Fixed N transferred to neighbouring reference plants modifies the N-15 signature of the soil N available both to the reference plant and the N-2-fixing legume. This provides strong support for using reference plants growing in mixture with the legumes for reliable quantifications of N-2 fixation.