Mechanisms for the retention of inorganic N in acidic forest soils of southern China

The mechanisms underlying the retention of inorganic N in acidic forest soils in southern China are not well understood. Here, we simultaneously quantified the gross N transformation rates of various subtropical acidic forest soils located in southern China (southern soil) and those of temperate forest soils located in northern China (northern soil). We found that acidic southern soils had significantly higher gross rates of N mineralization and significantly higher turnover rates but a much greater capacity for retaining inorganic N than northern soils. The rates of autotrophic nitrification and NH3 volatilization in acidic southern soils were significantly lower due to low soil pH. Meanwhile, the relatively higher rates of NO3− immobilization into organic N in southern soils can counteract the effects of leaching, runoff, and denitrification. Taken together, these processes are responsible for the N enrichment of the humid subtropical forest soils in southern China.


Separation of NH 4 + and NO 3 -
For isotopic analysis, NH 4 + and NO 3 -were separated by distillation with magnesium oxide and Devarda's alloy 1

Comparison of dilution and 15 N tracing model methods
To verify the utility of the 15 N tracing model for simulating the gross rates of N transformation that occur simultaneously, we compared the results obtained using the 15 N dilution method 3 with those obtained using the 15 N tracing method employed in this study.
The gross rates of mineralization, nitrate production, and NO 3 consumption per time interval and the average of all time intervals were calculated using the method of Kirkam and Bathalomew (1954) 3 . The results showed the calculated (dilution technique) and modelled ( 15 N tracing technique) mineralization and NO 3 production rates were similar (p < 0.01; Fig. 2A, B). Based on the generally strong agreement between the values, we were confident enough to use the modelled rates, which provided details on the N transformation mechanisms and therefore a much more detailed view of actual N dynamics. Therefore, the modelled rates were used to compare soil N transformations between temperate and subtropical-tropical forest soils in the present investigation.
Samples 14 and 15 were from the same core, but the 15 N tracing experiment was carried out in the laboratory for soil 14 (added 50 μg N g -1 soil) and in the field for soil 15 (added 2 μg N g -1 soil). The results showed that the N transformation rates determined in the laboratory and in the field were comparable ( Table 2 in article), and thus the N transformations were not stimulated in the laboratory due to higher N application in the studied acid forest soils. The laboratory studies can provide essential information to mechanistically understand the observed N cycling process in the field 4,5 , despite some reported problems 6 . Nitrogen enrichment in humid subtropical soils could therefore be explained using data obtained by comparing the gross rates of N transformation in laboratory-incubated soils with those of temperate soils ( Fig. 3 in article).
Previous studies, where the 15 N tracing approach (section 2.4 in article) was used, showed that N consumption/production rates can be separated into process-specific gross rates to provide more details about soil N transformations, despite the fact that there is still some uncertainty about the general nature of the application of the 15 N tracing model 5,7,8,9,10,11 . It should be stressed that, in general, individual gross N transformation rates cannot be directly measured but can only be quantified with the help of analytic models, the most basic of which was presented by Kirkham and Bartholomew (1954)  comparable results for mineralization and NO 3 production rates (p < 0.01; Fig. 1A, B). The advantage of the 15 N tracing approach is that not only pool-specific gross N rates, but also individual N rates, can be determined. Furthermore, 15 N tracing models overcome the restrictions of zero-order kinetic rates that are assumed in the dilution approach. First-order and Michaelis-Menten kinetics are more realistic because they take into account a non-linear behaviour with respect to changing N concentrations.
Discrepancies between the two methods may be associated with the use of different kinetics 11 (see also the result for NO 3 immobilization for the two neutral temperate forest soils [soil 7, 8], Fig. 1C).      Fig. 1 Comparison of modelled and calculated total mineralization (A), nitrification (B) and NO 3 immobilization rates (C; mg N kg -1 d -1 )

The results of analysis using average of soil properties and gross N transformation rates at each site
The modelled rates were simulated using the 15 N tracing model   10 , and the calculated rates were obtained using the method of Kirkam and Bathalomew (1954) 7 .