『Abstract
　Experiments with two well-studied denitrification and one recently isolated marine suboxic zone denitrifier show that the cellular-level denitrification N isotope effect (¹⁵ε) is typically lower than the canonical value of 〜25‰ under many conditions prevalent in the ocean. Across all three strains ¹⁵ε is 10-15‰ at cellular nitrate reduction rates that are more representative of the environment than the very high rates under which we and previous investigators measure ¹⁵ε to be 20-30‰. A sharp decrease in ¹⁵ε is also observed in individual nitrate drawdown assays as the extracellular nitrate concentrations approach 2-35μM and nitrate uptake becomes the rate-limiting step. On an apparently strain-specific basis, lower values of ¹⁵ε are observed under diverse conditions common in the natural environment: less reduced carbon sources, small inputs of oxygen, nutrient availability, agitation, and age of starter culture (i.e., initial of assays with cells that had recently depleted a large previous nitrate amendment or were more recently in the exponential growth (“bloom”) phase). A conserved oxygen-to-nitrogen isotope relationship across the experiments for all three denitrifiers (¹⁸ε/¹⁵ε=0.93±-.06 (1SD)) supports the interpretation that fractionation is imparted solely by the internal respiratory nitrate reductase, with the amplitude of ¹⁵ε varying with the proportional importance of cellular nitrate efflux relative to uptake. Aspects of the ¹⁵ε variation are unexpected; nevertheless, the occurrence of lower ¹⁵ε is robust. It is uncertain if our lower ¹⁵ε estimates apply to oceanic water column denitrification because field studies have generally yielded ¹⁵εwc between 20-30‰, more similar to previous culture estimates and our estimates at high cell specific nitrate reduction rates. If denitrification in the ocean's major suboxic zones does have an ¹⁵ε of 〜10-15‰, it would remove an apparent imbalance between global ocean N inputs and outputs previously suggested by fixed N isotope budgeting.』

1. Introduction
2. Methods
　2.1. Strains
　2.2. Denitrification assays
　　2.2.1. Preparation of starter cultures
　　2.2.2. Assay (re-suspension) medium composition
　　2.2.3. Reactor assembly and assay conditions
　　2.2.4. Species-specific denitrification assays
　　　2.2.4.1. Nitrate re-amendments for P. denitrificans
　　　2.2.4.2. Impact of inoculum age for P. chlororaphis
　　　2.2.4.3. Impact of mechanical stirring for Marinobacter sp.
　2.3. Sample processing
　2.4. Nitrate concentration and isotopic analyses
3. Results
4. Discussion
　4.1. Efflux model for respiratory nitrate reduction
　4.2. Cell specific nitrate reduction (CSNR)
　　4.2.1. Energy dependence of nitrate uptake, efflux and reduction
　　4.2.2. Influence of CSNR rate on ¹⁵εNaR
　4.3. Dependence of ¹⁵ε on culture conditions
　　4.3.1. Nitrate concentration and uptake limitation
　　4.3.2. Carbon and nutrient conditions
　　4.3.3. Oxygen
　　4.3.4. Growth phase of denitrifiers
　　4.3.5. Turbulence
　　4.3.6. Summary
　4.4. Implications for open ocean denitrification
　4.5. Global marine N isotope budget
5. Conclusions
Acknowledgments
Appendix A. Supplementary data
References