『Abstract
　Evidence for abiotic immobilization of nitrogen (N) in soil is accumulating, but remains controversial. Identifying the fate of N from atmospheric deposition is important for understanding the N cycle of forest ecosystems. we studied soils of two Abies pinsapo fir forests under Mediterranean climate seasonality in southern Spain - one with low N availability and the other with symptoms of N saturation. We hypothesized that biotic and abiotic immobilization of nitrate (NO3^-) would be lower in soils under these forests compared to more mesic temperate forests, and that the N saturated stand would have the lowest rates of NO3^- immobilization. Live and autoclaved soils were incubated with added ¹⁵NO3^- (10μg N g^-1 dry soil; 99％ enriched) for 24 h, and the label was recovered as total dissolved-N, NO3^-, ammonium (NH4⁺), or dissolved organic-N (DON). To evaluate concerns about possible iron interference in analysis of NO3^- concentrations, both flow injection analysis (FIA) and ion chromatography (IC) were applied to water extracts, soluble iron was measured in both water and salt extracts, and standard additions of NO3^- to salt extracts were analyzed. Good agreement between FIA and IC analysis, low concentrations of soluble Fe, and 100％ (±3％) recovery of NO3^- standard additions all pointed to absence of an interference problem for NO3^- quantification. On average, 85％ of the added ¹⁵NO3^- label was recovered as ¹⁵NO3^-, which supports our hypothesis that rates of immobilization were generally low in these soils. A small amount (mean = 0.06μg N g^-1 dry soil) was recovered as ¹⁵NH4⁺ in live soils and none in sterilized soils. Mean recovery as DO¹⁵N ranged from 0.6 to 1.5μg N g^-1 dry soil, with no statistically significant effect of sterilization or soil type, indicating that this was an abiotic process that occurred at similar rates in both soils. These results demonstrate a detectable, but modest rate of abiotic immobilization of NO3^- to DON, supporting our first hypothesis. These mineral soils may not have adequate carbon availability to support the regeneration of reducing microsites needed for high rates of NO3^- reduction. Our second hypothesis regarding lower expected abiotic immobilization in soils from the N-saturated site was not supported. The rates of N deposition in this region may not be high enough to have swamped the capacity for soil NO3^- immobilization, even in the stand showing some symptoms of N saturation. A growing body of evidence suggests that soil abiotic NO3^- immobilization is common, but that rates are influenced by a combination of factors, including the presence of plentiful available carbon, reduced minerals in anaerobic microsites and adequate NO3^- supply.

Keywords: Ferrous-wheel hypothesis; Retention; Iron; ¹⁵N; Nitrogen saturation status; Nitrogen cycle』

Introduction
Material and methods
　Site description
　Soil collection, preparation and sterilization
　¹⁵NO3^- amendment
　¹⁵N pool enrichment determination
　Statistical analysis
Results
　Fate of inorganic ¹⁵N added to soil
　Nitrate quantification
Discussion
　Methodological concerns
　Fate of ¹⁵N-NO3 added to soil
　Factors affecting abiotic retention of NO3^-
Acknowledgments
References