Denitrification refers to the reduction of the N oxides nitrate (NO₃^-) and nitrite (NO₂^-) to the gases nitric oxide (NO), nitrous oxide (N₂O) and dinitrogen (N₂). The process is generally considered to be carried out by facultative anaerobes, e.g. organisms that normally use oxygen (O₂₎ for respiration but in its absence use N oxides as electron acceptors. There is evidence that some organisms can simultaneously respire O₂ and NO₃^- in what has been termed "aerobic denitrification". Most denitrifying bacteria are heterotrophs, using organic carbon compounds as a source of energy. Bacteria capable of carrying out denitrification are widely distributed among world soils.

Denitrification is important in soil ecosystems for several reasons. First, removal of inorganic N by denitrification can influence the productivity of plants, since their growth is frequently limited by N. Second, denitrification is important to water quality. Nitrate is a federally listed drinking water pollutant and is an agent of eutrophication in marine ecosystems. Denitrification in soils, wetlands, streams and groundwater can prevent movement of NO₃^- from intensive upland land uses into aquatic ecosystems. Third, denitrification influences atmospheric chemistry through the production of N₂O. This gas contributes to the "greenhouse" effect and plays a role in ozone destruction. Finally, anaerobic metabolism is responsible for a significant portion of energy flow in many soils and wetlands. Denitrification is the most energetically favorable form of anaerobic metabolism, allowing for rates of energy generation close to those in aerobic metabolism and thus is essential to the overall microbial function of anaerobic (e.g. wet) soils.

While denitrification has been studied for over 100 years and the physiology of the process is relatively well understood, its importance in ecosystem, landscape and regional scale processes is still highly uncertain. The reason for this uncertainty is that denitrification is a difficult process to measure. Methods for measuring denitrification are flawed due either to alteration of substrate concentrations (either increasing or decreasing), disturbance of the physical environment, low sensitivity, or time and cost constraints. Quantification of denitrification has also been hindered by high spatial and temporal variation in activity. This variation is especially problematic given the lack of methods amenable to collection of large numbers of samples with reasonable expenditures of time and money.

We suggest that in addition to hindering quantification of denitrification in ecosystems, methodological problems have led to a systematic underestimation of N loss due to this process in temperate forest ecosystems. Denitrification is considered to be a very minor flux in the N budget of most natural terrestrial ecosystems and is considered to be an important sink for N in only a small group of agricultural production systems. However, we suggest that the dominant use of the C₂H₂ inhibition technique (described below) to measure denitrification has lead to serious underestimation of the importance of this process in these ecosystems. The use of C₂H₂ is problematic because it inhibits the supply of NO₃^- via nitrification and can therefore lead to NO₃^- limitation of denitrification. Inhibition of NO₃^- supply via nitrification is especially important in forest soils where NO₃ ^- concentrations are low, and denitrification has frequently been found to be limited by an absence of NO₃ ^-. Recent studies showing that gross rates of nitrification and NO₃^- immobilization in soil are much greater than previously thought suggest that the NO₃^- available to denitrifiers under natural conditions is much greater than measurement of ambient soil NO₃^- pools or rates of potential net nitrification would suggest. Inhibition of the vigorous turnover of NO₃^- by C₂H₂ could thus lead to systematic underestimation of denitrification rates in temperate forest soils and in many other ecosystems. Moreover, inhibition of general NO₃^- cycling in soils restricts the C₂H₂ inhibition method to measurement of denitrification in restricted activity centers, e.g. hotspots, potentially leading to an artificial increase in the variability of the process.

We have completed construction of a gas recirculation system for soil cores (Figure 1) and are in the process of making the preliminary tests to develop the measurement protocol. The design of the system follows the description in the proposal fairly closely.

We have completed the first series of experiments to determine flushing time for soil cores. These experiments involved overnight incubation of soil cores with SF₆ introduced into the atmosphere to allow diffusion of this gas into the micropores of the soil. Soils were then flushed with argon gas and measurements of SF₆ were made hourly. Conclusions from these experiments suggest that 3 hours is sufficient to flush gases from miropores (Figure 2).

Our central hypothesis concerning measurement of field rates of denitrification is that underestimation by acetylene based measurements is directly proportional to rates of nitrification. We have conducted experiments to compare measurements of N mineralization and nitrification via field and laboratory assays. This study was conducted at the Cary Arboretum in Millbrook NY and at several sites in the Catskill mountains. We measured net mineralization and net nitrification in the field using the static core method. Net mineralization and net nitrification were also determined in the laboratory on homogenized soil samples. Nitrification potential was determined in a soil slurry. Gross rates of mineralization as well as immobilization were determined using the ¹⁵N isotope dilution technique.

Both the net nitrification assay and the nitrification potential assay showed that nitrification in oak soils and maple soils that had high worm activity was close to 0 (Figure 3). Maple and Beech sites in the Catskills where there was no worm activity had high rates of nitrification. The nitrification potential assay was very sensitive at determining differences between sites.

Remaining work

We are now ready to begin denitrification measurements to evaluate the improved method with respect to measurement using acetylene. This evaluation will be done during the 1999 field season. We will also test the hypothesis that underestimation is directly proportional to the rate of nitrification using the nitrification gradient in the Catskill mountains that we identified last year.