Tansley Review No. 95 15 N natural abundance in soil‐plant systems

summary Equilibrium and kinetic isotope fractionations during incomplete reactions result in minute differences in the ratio between the two stable X isotopes, 15 N and 14 N, in various N pools. In ecosystems such variations (usually expressed in per mil [δ 15 N] deviations from the standard atmospheric N 2 ) depend on isotopic signatures of inputs and outputs, the input‐output balance, N transformations and their specific isotope effects, and compartmentation of N within the system. Products along a sequence of reactions, e.g. the N mineralization‐N uptake pathway, should, if fractionation factors were equal for the different reactions, become progressively depleted. However, fractionation factors van. For example, because nitrification discriminates against 15 N in the substrate more than does N mineralization, NH 4 + can become isotopically heavier than the organic N from which it is derived. Levels of isotopic enrichment depend dynamically on the stoichiometry of reactions, as well as on specific abiotic and biotic conditions. Thus, the δ 15 N of a specific N pool is not a constant, and 15 N of a N compound added to the system is not a conservative, unchanging tracer. This fact, together with analytical problems of measuring 15 N in small and dynamic pools of N in the soil‐plant system, and the complexity of the X cycle itself (for instance the abundance of reversible reactions), limit the possibilities of making inferences based on observations of 15 N abundance in one or a few pools of N in a system. Nevertheless, measurements of δ 15 N might offer the advantage of giving insights into the N cycle without disturbing the system by adding 15 N tracer. Such attempts require, however, that the complex factors affecting 15 N in plants be taken into account, viz. (i) the source(s) of N (soil, precipitation, NO X , NH 3 , N 2 ‐fixation), (ii) the depth(s) in soil from which N is taken up, (iii) the form(s) of soil‐N used (organic N, NH 4 + , NO 3 − ), (iv) influences of mycorrhizal symbioses and fractionations during and after N uptake by plants, and (v) interactions between these factors and plant phenology. Because of this complexity, data on δ 15 N can only be used alone when certain requirements are met, e.g. when a clearly discrete N source in terms of amount and isotopic signature is studied. For example, it is recommended that N in non‐N 2 ‐fixing species should differ more than 5% from N derived by N 2 ‐fixation, and that several non‐N 2 ‐fixing references are used, when data on δ 15 N are used to estimate N a ‐fixation in poorly described ecosystems. As well as giving information on N source effects, δ 15 N can give insights into N cycle rates. For example, high levels of N deposition onto previously N‐limited systems leads to increased nitrification, which produces 15 N‐enriched NH 4 and N‐depleted NO 3 . As many forest plants prefer NH 4 − they become enriched in 15 N in such circumstances. This change in plant 15 N will subsequently also occur in the soil surface horizon after litter‐fall, and might be a useful indicator of N saturation, especially since there is usually an increase in 15 N with depth in soils of N‐limited forests. Generally, interpretation of 15 N measurements requires additional independent data and modelling, and benefits from a controlled experimental setting. Modelling will be greatly assisted by the development of methods to measure the 15 N of small dynamic pools of N in soils. Direct comparisons with parallel low tracer level 15 N studies will be necessary to further develop the interpretation of variations in 15 N in soil‐plant systems. Another promising approach is to study ratios of 15 N: 14 N together with other pairs of stable isotopes, e.g. 13 C: 12 C or 18 O: 16 O, in the same ion or molecules. This approach can help to tackle the challenge of distinguishing isotopic source effects from fractionations within the system studied. CONTENTS Summary 179 I. Introduction 180 II. Units, causes of isotope effects, stoichiometry, modelling 181 III. N dynamics and variations in 15 N abundance in soil‐plant systems 183 IV. Applications 189 V. Conclusions and suggestions for future research 197 Acknowledgements 198 References 198