Enhancement of anammox by the excretion of diel SEE COMMENTARY vertical migrators Daniele Bianchia,b,1, Andrew R. Babbinc, and Eric D. Galbraitha aDepartment of Earth and Planetary Science, McGill University, Montreal, QC, Canada H3A 0E8; bSchool of Oceanography, University of Washington, Seattle, WA 98195; and cDepartment of Geosciences, Princeton University, Princeton, NJ 08544 Edited by David M. Karl, University of Hawaii, Honolulu, HI, and approved September 15, 2014 (received for review June 10, 2014) Measurements show that anaerobic ammonium oxidation with However, in situ measurements of denitrification and anam- nitrite (anammox) is a major pathway of fixed nitrogen removal in mox in anoxic waters have in most cases failed to show a close the anoxic zones of the open ocean. Anammox requires a source of coupling between these two biogeochemical pathways. While ammonium, which under anoxic conditions could be supplied by the measurements of fixed N removal with isotopically labeled sub- breakdown of sinking organic matter via heterotrophic denitrifica- strates show anammox occurring in many OMZs, denitrification tion. However, at many locations where anammox is measured, appears to be patchier (11). In incubation experiments, anammox + denitrification rates are small or undetectable. Alternative sources of proceeds as soon as labeled NH4 is added, while denitrification ammonium have been proposed to explain this paradox, for example can sometimes lag by multiple days following the addition of through dissimilatory reduction of nitrate to ammonium and trans- organic matter (e.g., ref. 12). This suggests that anammox could port from anoxic sediments. However, the relevance of these sources proceed at low, uniform rates in anoxic waters, while deni- in open-ocean anoxic zones is debated. Here, we bring to attention trification could be intermittent in space and time, fueled by high an additional source of ammonium, namely, the daytime excretion organic matter input events (13). It is possible that, despite the by zooplankton and micronekton migrating from the surface to local mismatch of observed rates, a stoichiometric balance be- anoxic waters. We use a synthesis of acoustic data to show that, tween anammox and denitrification is realized as an average on where anoxic waters occur within the water column, most migrators large spatial and temporal scales (11). However, the current pool spend the daytime within them. Although migrators export only of measurements is not extensive enough to test this hypothesis by a small fraction of primary production from the surface, they focus ex- providing an integrated view of fixed N removal in the water col- cretion within a confined depth range of anoxic water where particle umn, and the observed decoupling between anammox and deni- input is small. Using a simple biogeochemical model, we suggest trification remains puzzling due to the cryptic ammonium source. that, at those depths, the source of ammonium from organisms + Alternative sources of NH4 , other than the byproduct of undergoing diel vertical migrations could exceed the release from DNRN and denitrification, could be responsible for some of this particle remineralization, enhancing in situ anammox rates. The + decoupling. Different biogeochemical pathways can supply NH4 contribution of this previously overlooked process, and the numer- to anammox bacterial communities. These include dissimilatory ous uncertainties surrounding it, call for further efforts to evaluate nitrate reduction to ammonium (DNRA) (14), as well as diffusion the role of animals in oxygen minimum zone biogeochemistry. and advection from sulfate-reducing sediments, or from micro- aerobic remineralization at the boundaries of anoxic waters. anammox | denitrification | oxygen minimum zone | diel vertical migration However, it is not clear whether DNRA, thermodynamically dis- favored compared with DNRN and denitrification, is important in ater column oxygen minimum zones (OMZs), where oxy- anoxic open waters, compared with sediments (15). Similarly, + Wgen concentrations plummet to submicromolar levels (1), advection and diffusion of NH4 can be relevant at the boundaries are responsible for approximately one-third of the total removal of anoxic layers, or where anoxic waters impinge on sulfidic sedi- of fixed nitrogen from the oceans (2, 3). Several processes me- ments (8), but are of limited importance in the anoxic cores of EARTH, ATMOSPHERIC, diated by specialized prokaryotes convert fixed inorganic nitro- AND PLANETARY SCIENCES + − − gen (NH4 ,NO2 , and NO3 )toN2 in anoxic waters. Canonical − Significance denitrification, consisting of dissimilatory NO3 reduction to − NO2 (DNRN) followed by the further oxidation of organic − Nitrogen, the limiting nutrient for primary production across matter with NO (the denitrification step), was long considered 2 much of the ocean, is converted to biologically inactive N2 by the dominant fixed N removal pathway in anoxic waters. Over denitrification and anaerobic ammonium oxidation (anammox) the last decade, anammox has gained attention as a major sink of in anoxic waters. Anammox requires an active source of am- SCIENCES −3 < · ENVIRONMENTAL fixed N in nearly anoxic waters (O2 10 mmol m ) (4, 5). monium, which can be provided by concurrent denitrification. Stoichiometric considerations would suggest a close coupling However, anammox has been observed at high rates in the between denitrification and anammox (6, 7). Under anoxic + absence of denitrification, and the source of ammonium has conditions, the NH4 liberated by the remineralization of organic remained cryptic. Using a combination of observations and matter through DNRN and denitrification should accumulate in models, we suggest that zooplankton and micronekton pro- the water column because conventional (aerobic) nitrification vide a missing source of ammonium to anoxic waters through cannot proceed. However, this accumulation is not observed in diel vertical migrations, fueling anammox and decoupling it + the cores of anoxic waters, where observed NH4 concentrations from denitrification. This previously overlooked mechanism −3 are generally much less than 1 mmol·m (8). In these regions, can help to reconcile observations with theory and highlights + − oxidation of NH4 with NO2 by anammox is thought to be the the role of animals on ocean biogeochemistry. + + major sink of NH4 (9). Given that no significant NH4 accu- + mulation takes place in the cores of OMZs, the supply of NH4 Author contributions: D.B., A.R.B., and E.D.G. designed research; D.B., A.R.B., and E.D.G. by denitrification must be completely consumed by its consump- performed research; D.B. analyzed data; and D.B., A.R.B., and E.D.G. wrote the paper. tion via anammox. This would lead to ∼70% of the total fixed The authors declare no conflict of interest. nitrogen (N) removal occurring through denitrification, and This article is a PNAS Direct Submission. about 30% from anammox, with some variability related to the See Commentary on page 15604. stoichiometry of the organic matter respired (7, 10). 1To whom correspondence should be addressed. Email: [email protected]. www.pnas.org/cgi/doi/10.1073/pnas.1410790111 PNAS | November 4, 2014 | vol. 111 | no. 44 | 15653–15658 Downloaded by guest on September 23, 2021 OMZs, which are characterized by weak circulation and mixing animal communities (22–24). Our technique identifies the depth + (16, 17), small NH4 gradients (8), and undetectable oxygen con- of the main DVM stationary depth, at which the strongest sub- centrations over water layers hundreds of meters thick (1). euphotic sound scattering layer is centered during the daytime. + Here, we suggest that the excretion of NH4 by diel vertically The data that we analyzed cover the three major open-ocean migrating animals could play an important role in the N cycle of oxygen minima that host anoxic waters: the Northern Indian OMZs, and contribute to the decoupling of anammox and deni- Ocean (Arabian Sea), the Eastern Tropical North Pacific (ETNP), trification. As shown in DVM Association With Anoxic Waters,diel and the Eastern Tropical South Pacific (ETSP) OMZs. When vertical migrations (DVM) of zooplankton and micronekton plotted against in situ oxygen profiles from these regions, the commonly reach the anoxic layers of the major OMZs of the open acoustic data show that migrators enter waters with the lowest + ocean. Because animals excrete reduced N mostly as NH4 , they measured oxygen concentrations (Fig. 1). This picture is broadly can fuel anammox directly, and decouple it from denitrification. consistent with in situ net data, for example from the Arabian Sea Here, we present a first assessment of this process. and the ETNP, indicating diel migration of zooplankton and The rest of the paper is organized as follows. In DVM Asso- micronekton, including abundant euphausiids and myctophid ciation With Anoxic Waters, we show the common occurrence of fish, in and out of the oxygen-deficient layers (25–27). DVM across major OMZs by using acoustic data. In DVM Effects We quantified the overlap of DVM and anoxic waters by on the N Cycle in Anoxic Waters, we describe how the excretion of mapping the minimum O2 concentration observed at the depth + NH4 by vertical migrators can alter the balance between fixed N of migration, as shown in Fig. 2. Where available, we used the removal pathways, enhancing anammox. In Spatial Models of DVM direct DVM depth estimates from acoustic data from ref. 21. Amplification