Davidson et al. 1991

Davidson EA, Hart SC, Shanks CA, Firestone MK. 1991. Measuring gross nitrogen mineralization, immobilization, and nitrification by 15N isotopic pool dilution in intact soil cores. Journal of Soil Science 42: 335-349.

These authors evaluated the use and limitations of the isotope-pool dilution technique when studying nitrogen dynamics in soil. Because addition of inorganic nitrogen compounds (NH4+, NO3-) can stimulate microbial activity in N-limited systems such as most soils, estimating the rate of these processes by tracking 15N through a system will almost certainly overestimate these rates. The isotope-pool dilution method, on the other hand, measures the dilution of enrichment in the nitrogen pool at the end of a particular process, relying on the assumption that additional product of metabolism will have negligible effects on the magnitude of that metabolism. In this study, immobilization of nitrogen was the main focus of investigation, comparing 15N isotope dilution in pools of either 15NH4+ or 15NO3-.

There are three key assumptions for the isotope-pool dilution method in this context. 1. Microorganisms do not discriminate between 15N and 14N; 2. rates of processes measured remain constant over the incubation period; 3. 15N assimilated during the incubation period is not remineralized. Previously, these assumptions had been evaluated for well-mixed soils, but not for unmixed field-collected soil samples. While fractionation by biological processes certainly does result in discrimination between isotopes of nitrogen, it is of negligible importance when injected solutions are very highly enriched and incubation periods are relatively short; in this case, injections were more than 90% 15N and incubations ran for 24 hours. Rates of measured processes will change if the population and / or activity of microorganisms changes, but again, over a 24-hour incubation period under controlled conditions this is unlikely. Highly enriched injections allow the use of small injection volumes, limiting the impact of nutrient enrichment. These authors were able to measure the remineralization of immobilized 15N, and estimated that between 1.0 and 1.6% of injected 15NH4+ appeared in the 15NO3- pool after 24 hours; they consider this an insignificant amount, but caution that longer incubations would almost certainly result in much more problematic amounts of remineralization.

This paper is clearly a major part of the basis of the project I am currently engaged in with Katherine from our 2009 field season at Alexandra Fjord. I probably should have read this paper long ago. The three main conclusions stated by these authors at the end of their paper I think can be quoted verbatim as justification for both why I (should have earlier) read this paper, and as a reminder to myself to include this paper in the methods & materials section of the eventual manuscript.

"Three points should be considered when applying the isotope dilution method.
1. Accurate estimation of both 14N and 15N initial pool sizes is important. Abiotic consumption of label, such as by clay fixation, can cause significant errors. A subset of intact cores may need to be destructively sampled directly after adding 15N for estimation of initial pool sizes.
2. Homogeneity of 15N enrichment throughout a soil sample is not possible, and perfectly uniform distribution of added label is not necessary. However, significant errors can arise from a bias in 15N distribution that is concurrent with a non-random
distribution of microbial processes. Distribution of label should, therefore, be as uniform as possible.
3. In situ gross immobilization rates may be overestimated by isotope dilution methods and underestimated by chloroform fumigation methods, depending on which (if any) kN factor is applied to the latter. Gross mineralization and gross nitrification estimates from isotope dilution are more reliable because these rates should not be affected by addition of 15N label in the form of the process products."

Miller et al. 2008

Miller MN, Zebarth BJ, Dandie CE, Burton DL, Goyer C, Trevors JT. 2008. Crop residue influence on denitrification, N2O emissions and denitrifier community abundance in soil. Soil Biology & Biochemistry 40: 2553-2562.

These authors conducted a factorial experiment using packed soil cores to examine the influence of varying levels of available carbon and nitrogen on the process of denitrification. The treatments consisted of addition of glucose at three levels and KNO3 at four levels in experiment 1, and additions of either red clover or barley straw crop residues with or without additional KNO3. They measured soil chemistry, including extractable organic carbon and NO3- concentration based on K2SO4 extractions, as well as N2O production, the molar ratio of N2O (N2O : (N2O + N2)), and a handful of bacterial genes by qPCR.

The experimental setup was very similar to what we used in SLSC 802 (Special Topics) in the fall of 2010; cylindrical soil cores with gas-exchange holes in the sides were filled with soil at 1 g cm-3 bulk density and a water content of 70% and placed in 1 L canning jars with lids fitted with a perforable septum for gas sampling. One important difference between this experiment and what we have done is that in this experiment, gas measurements were of total cumulative gas production, whereas we flushed each jar with ambient air after each sampling event. Presumably this difference will have important effects on the formation of anaerobic conditions and microbial consumption of N2O previously produced under less anoxic conditions.

Not surprisingly, minimal denitrification activity was found in treatments without added NO3-. Starting NO3- concentrations were 3 mg NO3--N kg-1 soil, and fell in all treatments without added NO3- to less than 1 mg NO3--N kg-1. Once this supply of readily available nitrate was used, it appears the bacteria ceased denitrification activity, or at least it was reduced.
The red clover had a much lower C:N ratio than the barley straw, 13:1 and 45:1, respectively, and more labile carbon. This difference appears to have driven the observed difference in denitrification activity, in a manner that reflects the results of the simple-C-source experiment 1. In general, more labile C and more available N leads to stronger denitrification activity and greater production of N2O; in sealed jars such as these, strong respiration under these conditions leads to anaerobic conditions and a fall in the molar ratio of N2O as nosZ-equipped microbes consume N2O as a terminal electron acceptor.

Extractable organic carbon (EOC) was a relatively poor predictor of denitrification, compared to respiration as measured by CO2 production. EOC is a measure of the instantaneous size of the pool of labile C, while respiration represents carbon that has already passed through a microbe’s metabolism. The distinction here may be between two different pools of carbon, as well as between an instantaneous snapshot measure and a series of measurements readily convertible to an estimate of the rate of a process.

In conclusion, these authors reiterate their finding that available C and available N (especially as NO3-) are strong predictors of denitrifying activity, across a range of C and N sources. I read this paper for the class SLSC 802 in the fall of 2010, but the portions describing denitrification physiology and especially the qPCR information will be generally useful to my other projects.