Widespread occurrence of nitrate storage and denitrification among Foraminifera and Gromiida

Abstract

Benthic foraminifers inhabit a wide range of aquatic environments including open marine, brackish, and freshwater environments. Here we show that several different and diverse foraminiferal groups (miliolids, rotaliids, textulariids) and Gromia, another taxon also belonging to Rhizaria, accumulate and respire nitrates through denitrification. The widespread occurrence among distantly related organisms suggests an ancient origin of the trait. The diverse metabolic capacity of these organisms, which enables them to respire with oxygen and nitrate and to sustain respiratory activity even when electron acceptors are absent from the environment, may be one of the reasons for their successful colonization of diverse marine sediment environments. The contribution of eukaryotes to the removal of fixed nitrogen by respiration may equal the importance of bacterial denitrification in ocean sediments.

Only two eukaryotes, the marine benthic foraminiferal species Globobulimina turgida* and Nonionella cf. stella (1, 2), are known to carry out complete denitrification of nitrate to N₂. They accumulate intracellular nitrate to millimolar concentrations, which they subsequently respire in the absence of oxygen. It has been uncertain whether this is a previously undescribed evolutionary trait found only in a few closely related species that share the same mode of life or an old or coevolved widespread trait in Foraminifera and other eukaryotes. Our finding of denitrification via nitrate pools in mobile eukaryotic organisms adds a previously undescribed dimension to the marine nitrogen cycle. Because benthic foraminifers inhabit a wide range of aquatic environments and can be found in densities of up to several million individuals per square meter (3), they may play an important role in temporal nitrate sequestering, nitrate transport, and nitrogen removal through denitrification.

G. turgida and Nonionella cf. stella belong to the foraminiferal order Rotaliida, inside which they group into different clades according to molecular phylogeny (4). Both taxa thrive in oxygen-free sediment environments (5–7) where alternative electron acceptors such as nitrate are required for respiration. Many other genera have been observed in such environments (e.g., Chilostomella, Stainforthia, Bolivina, Uvigerina, Bulimina, and Reophax) (8), and recently published data suggest that the use of nitrate is not confined to only two genera. In a Swedish fjord, nitrate detected in G. turgida accounted for only about 20% of the total cell-bound nitrate pool in the sediment (2), suggesting that other foraminifers might also accumulate nitrate. Stainforthia sp. from the oxygen minimum zone (OMZ) of the continental shelf off Chile (1) and Uvigerina akitaensis, Bolivina spissa, and Textularia sp. from Japanese deep-ocean margin sediment (9) also have been shown to accumulate nitrate.

To evaluate the evolutionary origin, environmental affiliation, and biogeochemical importance of denitrification in foraminifers, we have determined the denitrification rates of seven more species and measured the nitrate content of 67 foraminiferal species sampled from different marginal marine environments: Aiguillon Bay (France), Bay of Biscay (France), Disko Bay (Greenland), Gullmars Fjord (Sweden), Limfjorden (Denmark), the North Sea, the Peruvian–Chilean OMZ, the Rhône prodelta (France), and the Skagerrak. In addition, 55 specimens belonging to the genus Gromia were tested for nitrate content. This aquatic ameboid protist genus bears an organic test and is thought to be the sister group of foraminifers within Rhizaria (10).

Results and Discussion

Taxonomic Diversity Among Nitrate-Storing Foraminifers.

Up to 48 specimens were analyzed individually from each foraminiferal species to determine nitrate accumulation. A species was confirmed as a nitrate collector if the intracellular nitrate concentration was 100 μM or more than 10-fold the maximum concentration in the environment (Table S1). More than half of the tested foraminiferal species and all of the gromiids were confirmed as nitrate collectors (Table 1).

Table 1.

Intracellular nitrate content and concentration in foraminifers and gromiids

			NO3− (pmol per cell)
Species	Location	n	Mean (±SEM)	Range	Volume (mm3)*	NO3− (mM)*
Allogromiids
Agglutinated sp.	Rhône Delta	1	0	0		0
Bathysiphon cf. argenteus	OMZ-Perú	2	0	0		0
Bathysiphon minutus	Skagerrak	2	0	0		0
Crithionina hispida	OMZ-Perú	4	0	0		0
Hippocrepinella alba	Skagerrak	6	5 (1)	1–6	2.0E−01 (5.0E−02)	0
Komokiacea	OMZ-Perú	3	0	0		0
Pelosina variabilis	Skagerrak	3	50 (25)	6–92	2.2E−01 (2.2E+02)	0
Saccammina sp.	Bay of Biscay	2	0	0		0
Technitella legumen	Skagerrak	13	5 (1)	1–10	4.0E−01 (4.4E−02)	0
Miliolids
Biloculinella depressa	North Sea	1	0	0		0
Pyrgo elongata	Rhône Delta	9	43 (14)	19–139	4.7E−02 (5.8E−03)	0.8 (0.2)
Pyrgo williamsoni	North Sea	1	5	5	4.7E−02	0.10
Pyrgoella sphaera	North Sea	2	6 (1)	5–7	4.7E−02 (5.8E−03)	0.1 (0.02)
Quinqueloculina sp.	OMZ-Perú	2	0	0		0
Quinqueloculina seminulum	Skagerrak	2	0	0		0
Quinqueloculina seminulum	Bay of Biscay	10	0	0		0
Quinqueloculina seminulum	Rhône Delta	2	0	0		0
Triloculina tricarinata	North Sea	1	0	0		0
Lagenids
Dentalina sp.	Rhône Delta	1	0	0		0
Rotaliids
Unknown clade
Bolivinita quadrilatera	Bay of Biscay	1	0	0		0
Cancris inflatus	OMZ-Perú	18	262,877 (4,253)	3,920–76,475	1.2E−01 (2.4E−02)	262 (37)
Gyroidina altiformis	Bay of Biscay	1	0	0		0
Gyroidina neosoldanii	OMZ-Perú	6	13,190 (480)	25–3,375	2.7E−02 (1.2E−02)	241 (46)
Nonion scaphum	Rhône Delta	9	0	0		0
Nonion scaphum	Bay of Biscay	2	0	0		0
Valvulineria bradyana	Rhône Delta	16	1,268 (164)	176–2,541	1.5E−02 (1.4E−03)	95 (15)
Valvulineria cf. laevigata	OMZ-Perú	17	865 (640)	55–1,093	1.9E−02 (3.7E−03)	25 (12)
Clade 1
Bolivina alata	Bay of Biscay	6	615 (154)	188–1,266	1.7E−02 (1.1E−03)	37 (12)
Bolivina cf. abbreviata	OMZ-Perú	14	1,081 (368)	26–4,934	1.2E−02 (2.7E−03)	153 (49)
Bolivina cf. skagerrakensis	North Sea	1	83	83	1.7E−02	5
Bolivina plicata	OMZ-Perú	24	478 (72)	59–1,037	7.5E−03 (9.5E−04)	79 (15)
Bolivina seminuda	OMZ-Perú	10	564 (135)	187–807	5.2E−03 (1.8E−03)	118 (18)
Bolivina subaenariensis	Bay of Biscay	47	285 (46)	43–1,023	2.5E−02 (4.3E−03)	44 (9)
Cassidulina carinata	Rhône Delta	23	3(1)	0–21	4.1E−03 (2.2E−04)	1 (0.5)
Cassidulina cf. laevigata	North Sea	1	21	21	4.1E−03	5
Cassidulina cf. laevigata	OMZ-Perú	6	523 (289)	194–1,388	1.2E−02 (3.6E−03)	41 (12)
Globobulimina affinis	OMZ-Perú	10	1,298 (2,320)	784–24,853	5.7E−02 (1.0E−02)	310 (97)
Globobulimina affinis	Bay of Biscay	1	3,290	3,29,018	9.3E−02 (2.0E−02)	292
G. auriculata cf. arctica	Greenland	11	10,624 (3,555)	0–34,902	1.0E−01 (1.7E−02)	113 (43)
Globobulimina cf. ovula	OMZ-Perú	5	3,369 (1,602)	633–7,600	1.0E−03 (2.3E−03)	375 (174)
Globobulimina turgida	Gullmar fjord	20	18,000 (4,852)	0–72,000	5.0E−01 (3.6E+00)	10 (2)
Globobulimina turgida	Skagerrak	17	8,192 (1,497)	15–15,982	1.0E−01 (1.7E−02)	71 (13)
Uvigerina elongatastriata	Bay of Biscay	4	274 (244)	0–1,003	5.1E−03 (5.7E−04)	60 (55)
Uvigerina elongatastriata	Rhône Delta	2	5,389 (687)	4,703–6,077	2.4E−02 (9.4E−03)	286 (143)
Uvigerina mediterranea	Bay of Biscay	8	101 (66)	0–537	2.0E−02 (6.6E−03)	6 (4)
Uvigerina peregrina	Bay of Biscay	2	0	0		0
Uvigerina peregrina	North Sea	5	332 (184)	17–792	2.0E−02 (6.6E−03)	16 (9)
Uvigerina phlegeri	Rhône Delta	48	444 (44)	29–1,092	8.4E−03 (1.8E−04)	209 (48)
Clade 2
Ammonia beccarii	Rhône Delta	6	0	0		0
Ammonia beccarii	Bay of Biscay	9	0	0		0
Pseudoeponides falsobeccarii	Rhône Delta	1	0	0		0
Ammonia sp.	Limfjorden	14	1 (0.5)	0–5	2.3E−01 (1.7E−02)	0
Ammonia tepida	Aiguillon Bay	15	13 (4)	0–58	2.3E−01 (1.7E−02)	0
Haynesina germanica	Aiguillon Bay	13	0	0		0
Hyalinea balthica	North Sea	14	8 (2)	0–30	8.0E−03 (1.2E−01)	1 (0.3)
Clade 3
Bulimina aculeata	Rhône Delta	7	0	0–2	6.0E−03 (4.3E−04)	0
Bulimina aculeata	Bay of Biscay	6	19 (12)	0–63	7.4E−03 (3.8E−04)	3 (2)
Bulimina cf. elongata	OMZ-Perú	5	817 (287)	263–1,812	7.9E−03 (1.2E−03)	116 (43)
Bulimina marginata	Skagerrak	4	5 (2)	0–9	1.1E−03 (1.1E−02)	0.5 (0.2)
Bulimina marginata	Bay of Biscay	14	40 (4)	40–60	3.2E−02 (1.1E−03)	4 (1)
Bulimina marginata	Rhône Delta	1	0	0		0
Chilostomella oolina	Bay of Biscay	4	1,124 (520)	185–2,611	2.0E−02 (2.0E−03)	65 (36)
Cibicidoides pachyderma	Bay of Biscay	1	0	0		0
Epistominella exigua	OMZ-Perú	2	0	0		0
Melonis barleeanus	North Sea	7	9 (3)	1–27	1.4E−02 (2.0E−02)	0.6 (0.2)
Melonis barleeanus	Rhône Delta	2	0	0		0
Nonionella cf. stella	OMZ-Chile	43	186 (24)	8–794	5.2E−03 (7.1E−04)	35 (5)
Stainforthia sp. var. I	OMZ-Chile	26	60 (46)	0–172	3.3E−04 (2.1E−05)	180 (29)
Textulariids
Clavulina cylindrica	Bay of Biscay	6	1,941 (314)	568–2,651	3.7E−02 (5.8E−03)	61 (12)
Clavulina cylindrica	Rhône Delta	3	2,202 (480)	840–3,672	3.5E−02 (9.8E−04)	48 (13)
Cyprissubglobosus	Bay of Biscay	1	0	0		0
Cyclammina cancellata	OMZ-Perú	2	45,563 (45,563)	209–90,915	3.8E−01 (3.1E−03)	119 (118)
Cyclammina cancellata	Bay of Biscay	1	0	0		0
Goesella flintii	OMZ-Perú	6	459 (424)	0–2,575	1.0E−01 (2.7E−02)	24 (23)
Labrospira cf. kosterensis	OMZ-Perú	6	3,139 (845)	474–5,016	5.1E−02 (1.2E−02)	57 (12)
Labrospira cf. subglobosa	OMZ-Perú	2	0	0		0
Nouria polymorphinoides	Bay of Biscay	2	0	0		0
Pseudoclavulina crustata	Bay of Biscay	1	598	598.00	2.90E−01	2
Reophax micaceus	Bay of Biscay	3	0	0		0
Reophax sp.	OMZ-Perú	1	0	0		0
Rhabdammina inaequalis	North Sea	1	0	0		0
Textularia cf. tenuissima	OMZ-Perú	2	450 (432)	406–493	1.1E−02 (2.9E−03)	43 (7)
Other Rhizaria
Gromia sp.	North Sea	11	14,682 (4,649)	158–47,526	1.6E−01 (3.5E+00)	140 (46)
Gromia sp.	Skagerrak	12	35,277 (16,546)	158–2,04,000	5.1E−01 (1.1E−01)	53 (19)
Gromia sp.	OMZ-Perú	4	29,328 (7,589)	13,723–50,000	1.0E−01 (4.0E−02)	567 (283)
Gromia sp.	Bay of Biscay	2	2,846 (1,275)	1,571–4,121	9.3E−02 (2.0E−02)	35 (21)
Gromia sp.	Rhône Delta	13	3,889 (1,024)	576–12,676	1.6E−01 (1.1E−01)	91 (26)
Gromia sp.	Greenland	13	12,997 (2,954)	1,205–35,911	8.0E−02 (2.3E−02)	163 (54)

Species with a intracellular nitrate concentration > 0.1 mM are those confirmed as nitrate collectors. The taxonomic classification accommodates traditional and molecular phylogenetic systems: (i) allogromiids sensu Cedhagen et al. (11)—unilocular foraminifers with an organic or agglutinated test (traditional orders Allogromiida and Astrorhizida), (ii) miliolids—multilocular foraminifers with a calcitic porcelaneous test (order Miliolida), (iii) lagenids—multilocular foraminifers with a monolamellar perforated calcitic test (order Lagenida), (iv) rotaliids sensu lato—multilocular foraminifers with a bilamellar perforated calcitic test (orders Globigerinida, Rotaliida and Buliminida), and (v) textulariids sensu lato—multilocular foraminifers with an agglutinated test (orders Lituolida, Trochamminida and Textulariida).

*Values are mean values. SEM is given in parentheses.

Nitrate-collecting species were found within the miliolids, rotaliids, and textulariids. However, none of the six tested allogromiids or the single tested lagenid contained measurable nitrate. All of the tested gromiids contained nitrate with average internal concentrations ranging from 53 to 566 mM (Table 1).

Nitrate collection capacity was broadly distributed in genera within all three rotaliid clades proposed by Schweizer et al. (4) (Table 1): Bolivina, Cassidulina, Globobulimina, and Uvigerina (clade 1); Hyalinea (clade 2); Bulimina, Chilostomella, Melonis, Nonionella, and Stainforthia (clade 3); and also in genera not yet represented in phylogenetic analyses such as Cancris, Gyroidina, and Valvulineria. Within textulariids, nitrate collection was found in Cyclammina, Labrospira, Pseudoclavulina, Textularia, and members of the Valvulinidae family, and, in miliolids, Pyrgo elongata displayed the trait (Table 1).

Nitrate-collecting species within a genus showed variable nitrate content, and the magnitude of intracellular nitrate content generally did not map onto the foraminiferal phylogeny. The nitrate content seemed to reflect different physiological and environmental conditions because considerable intraspecific variation in intracellular nitrate concentration was observed among the species confirmed as nitrate collectors. The coefficient of variation was 48–160% for species where n > 10, and some individuals were completely void of nitrate. This intraspecific variation indicates that a dynamic intracellular nitrate pool is replenished and depleted, depending on the history of nitrate exposure, growth, starvation, and dormancy of the organisms. Exposure to oxygen might also lead to reduced nitrate content. Melonis barleeanus sampled in the oxic zone of the Rhône Delta sediment did not contain nitrate whereas the same species sampled in the nitrate reduction zone in the North Sea contained nitrate.

Among the species with at least 10 replicates measured, only Ammonia sp., Ammonia tepida, Haynesina germanica, Quinqueloculina seminulum, Nonion scaphum, and Technitella legumen showed a consistent absence of nitrate in their cells. We cannot conclude whether the lack of measurable nitrate accumulation results from the absence of a denitrification pathway or whether the species might accumulate nitrate under other environmental conditions. This applies even more to taxa that are poorly replicated in the present study (Biloculinella depressa, Bolivinita quadrilatera, Cibicidoides pachyderma, Cribrostomoides subglobosus, Dentalina sp., Rhabdamina inequalis, and Triloculina tricarinata; n = 1). Thus, nitrate storage could be even more widely distributed within Foraminifera than indicated by the present data set.

Nitrate Respiration.

In previous studies the ability of foraminifers to store intracellular nitrate was found to be associated with the ability of these organisms to respire nitrate through complete denitrification to N₂ (1, 2). Using a subset of the species analyzed for intracellular nitrate, we measured denitrification as N₂O production after acetylene inhibition of N₂O reduction to N₂. This method may underestimate but will never overestimate denitrification (12), and rates should therefore be considered as minimum estimates.

Denitrification capacity was found for all analyzed species that contained intracellular nitrate (Table 2), thus confirming that the intracellular nitrate pool was used for respiration. Denitrification rates ranged between 45 and 248 pmol nitrogen individual⁻¹ d⁻¹, which is close to the activities measured previously in G. turgida and Nonionella cf. stella (1, 2). There was no N₂O production associated with the nonnitrate-accumulating foraminifer A. tepida.

Table 2.

Denitrification and oxygen respiration rates of various foraminifers

Species	Individuals*	Denitrification (pmol nitrogen individual−1 d−1)	Individuals†	Oxygen respiration (pmol O2 individual−1 d−1)
Ammonia tepida (Aiguillon Bay)	2	0 (n = 1)	2	2030 ± 72 (n = 4)
Bolivina subaenariensis (B.Biscay)	10–12	78 ± 2 (n = 2)	8	252 ± 50 (n = 3)
Uvigerina phlegeri (Rhône)	10	46 ± 2 (n = 1)	5–8	90 ± 25 (n = 3)
Valvulineria bradyana (Rhône)	10	183 ± 10 (n = 2)	2–4	759 ± 288 (n = 3)
Nonionella cf. stella (OMZ, Chile)	3–5	84 ± 33 (n = 3)‡	3–5	760 ± 87 (n = 3)§
Globobulimina turgida (Gullmar fjord)	3	565‡
Bolivina plicata (OMZ, Perú)	3	79 (n = 1)
Valvulineria cf. laevigata (OMZ, Perú)	10	248 ± 180 (n = 2)	7	754 ± 146 (n = 2)
Bolivina seminuda (OMZ, Perú)	3	216 (n = 1)	3	368 (n = 1)
Stainforthia sp. (OMZ, Perú)	4	70 (n = 1)	4	822 (n = 1)

Rates are given as the mean (±SEM); n designates the number of replicates.

*Individuals in the measuring chamber during denitrification measurements.

^†Individuals in the measuring chamber during oxygen respiration measurements.

^‡Data are from ref. 2.

^§Data are from ref. 1.

Interestingly, four species collected in the Peruvian OMZ produced N₂O in the absence of acetylene (Bolivina plicata, Bolivina seminuda, Valvulineria cf. laevigata, Stainforthia sp.), suggesting a lack of nitrous oxide reductase in these organisms and thus making them greenhouse gas sources.

All tested denitrifying species could also respire with oxygen (Table 2)—even those collected in oxygen-free environments such as the Peruvian OMZ. The denitrifying foraminifers should therefore be regarded as facultative anaerobes. The oxygen respiration rates were generally about 3–13 times higher than the denitrification rates and, given the generally higher energy yield from oxic respiration (13), this could indicate that denitrification is an auxiliary metabolism used for cell maintenance, food collection, and locomotion during temporary stays in oxygen-free environments, whereas oxygen might be required for growth and reproduction.

Ecological Diversity of Nitrate-Storing Foraminifers.

Nitrate storage was common in foraminifers from very diverse benthic marine environments. In the OMZ, 16 of 23 tested species stored nitrate. More widespread species such as Uvigerina peregrina, Valvulineria bradyana, and Clavulina cylindrica from continental slopes, shelves, and coastal sediments (3) also store nitrate. Additionally, nitrate collectors were found in bathyal sediments (e.g., Bulimina marginata), and several of the species sampled in the Rhône Delta also accumulated nitrate. Species typically in intertidal mudflats such as A. tepida and Haynesina germanica did not store nitrate, and so far this environment does not seem to house nitrate-storing foraminifers (Table 1).

Nitrate collection capacity is also found for foraminifers occupying various microhabitats. Some of the nitrate-collecting taxa prefer oxygen-depleted environments such as the sediments within the OMZs (see Table 1 and ref. 8) or layers in the sediment where oxygen and even nitrate are absent from the pore water [e.g., the deep-living sedimental infaunal genera Globobulimina and Chilostomella (7, 8)]. Other nitrate-storing species, such as Bolivina subaenariensis and Uvigerina mediterranea, are opportunistic with respect to oxygen and may be found in both oxic and oxygen-free environments (5, 14, 15). Species that are mainly found in oxic microhabitats, such as Cassidulina carinata and P. elongata (16, 17), can, however, also store nitrate.

This almost ubiquitous presence of nitrate-collecting foraminifers suggests that the trait is one of the reasons for their successful colonization of marine sediments. As facultative anaerobes, the nitrate-storing foraminifers can use either oxygen or nitrate in the environment for respiration. The combination of nitrate storage and nitrate respiration enables the organisms to sustain respiratory activity in shorter or longer periods when oxygen or/and nitrate is absent from the environment. This allows them to explore food resources and to periodically seek shelter from predation in deeper sediment layers and also to survive exposure to environmental fluctuations such as passive transport into anoxic environments caused by sediment resuspension and macrofauna-mediated bioturbation events.

Molecular Phylogeny and Evolutionary Implications.

The molecular phylogeny of foraminifers based on partial small subunit (SSU) ribosomal DNA (rDNA) sequences inferred here (Fig. S1) is congruent with previous studies. Allogromiids form a paraphyletic group at the base of the tree, whereas miliolids and rotaliids/textulariids (together with allogromiids of clades A and C) form two separated clades arising from allogromiids (18, 19). At the moment, there is no published sequence of lagenids, but this group branches as a sister group of the textulariid/rotaliid clade according to previous results (20). The polyphyly of textulariids (18) as well as the monophyly of rotaliids sensu lato, with Buliminida and Globigerinida included (4, 21) can also been observed. There is a lack of taxonomic sampling to determine the molecular phylogenies of miliolids or textulariids. However, most of the morphological superfamilies of the rotaliids have been sampled, and for these we now have a good overview (Fig. S1 and ref. 22).

Nitrate-collecting rotaliids belong mainly to clades 1 and 3. Hyalinea balthica is the only nitrate-collecting foraminifer within clade 2, which primarily encompasses shallow water and epifaunal/planktic species. Nitrate collection was also found among the textulariids and miliolids and even among the gromiids (Fig. S1), which probably share a common ancestor with Foraminifera (10). Species that may lack the ability to accumulate nitrate are likewise represented within all major groups (Fig. S1), suggesting a complex evolutionary history of nitrate accumulation and denitrification within Foraminifera.

Assuming that none of the nitrate-collecting species depend on symbiotic bacteria for denitrification (2), two hypotheses arise concerning their evolutionary history. Either the trait appeared during the Neoproterozoic in a common ancestor of foraminifers and gromiids and was possibly followed by several losses in separate phylogenetic lineages or it was acquired several times independently in different lineages in more recent history. In other words, acquisition of denitrifying genes among foraminifers occurred either through the primitive endosymbiosis leading to denitrifying mitochondria or through subsequent lateral gene transfer(s). Denitrification relies on a large cluster of genes (23), and even when found in eukaryotic fungi, most of these genes apparently have a bacterial origin (24). If the same holds true for the foraminifers and gromiids, it is likely that denitrification was incorporated with the protomitochondrion in the very first eukaryotes and that more eukaryotic phyla could have retained the trait.

Nitrate-Storing Foraminifers: Implications for the Marine Nitrogen Cycle.

Prokaryotic denitrification and anaerobic ammonium oxidation are considered to be the only processes returning combined nitrogen to the atmosphere from the sea (25). The widespread occurrence of nitrate storage and denitrification among foraminifers demonstrated here indicates that eukaryotes may also play a role.

Total foraminifer-mediated denitrification in various environments can be estimated by combining the abundance of living nitrate-storing foraminifers and denitrification rates measured on isolated specimens, assuming that the rates estimated here are representative of the organisms in their natural habitat.

Corliss and Weering enumerated the foraminiferal population in the shelf sediments from the Skagerrak (26). Among the taxa that they found, three have a documented denitrification capacity: Bolivina sp., Globobulimina sp., and Uvigerina sp. (Table 2). Using the abundance data for these taxa at site A84-1 in ref. 26 and their cell-specific denitrification rate (Table 2), foraminiferal denitrification is about 720 μmol nitrogen m⁻²·d⁻¹. For comparison, the denitrification rate measured close to this site is 1030 μmol nitrogen m⁻²·d⁻¹ (site BB12 in ref. 27), suggesting that foraminifers are responsible for up to 70% of the measured benthic denitrification activity. Likewise, in a canyon of the Bay of Biscay, B. subaenariensis alone with 82 cells/cm² (15) would produce about 64 μmol N₂–nitrogen m⁻²·d⁻¹. Total denitrification estimated from the nitrate pore water profiles measured at the site (28) is around 76 μmol nitrogen m⁻²·d⁻¹, suggesting that foraminifers are quantitatively important denitrifiers at this location too. In part of the OMZ off Chile where foraminiferal abundance is high [205 cells cm⁻² (1)], foraminifer-mediated denitrification is 173 μmol N d⁻¹ (1). Total benthic denitrification reported for this site is around 250 μmol N m⁻²·d⁻¹ (29), indicating that foraminifers may account for almost 70% of the nitrogen loss from the sediment. In the Arabian Sea OMZ, however, the foraminifers contribute less to denitrification. With a density of 48 denitrifying foraminifers/cm² (30), foraminifer-based denitrification accounts for 78 μmol nitrogen m⁻²·d⁻¹, which corresponds to 9–15% of the measured benthic denitrification [510–840 μmol nitrogen m⁻²·d⁻¹ (31)]. In estuaries, foraminiferal denitrification may also play a significant role in the loss of combined nitrogen, and it may thereby mitigate coastal eutrophication. Using the standing stock of Uvigerina phlegeri, G. turgida, and Valvulineria bradyana in the Tagus prodelta (32), and their cell-specific denitrification rate, foraminiferal denitrification rates range between 72 and 240 μmol nitrogen m⁻²·d⁻¹. Denitrification measured there in July–October, which corresponds to the period when the above-mentioned species were enumerated (32), is 480–960 μmol nitrogen m⁻²·d⁻¹ (33). The foraminifers might thus contribute 8–50% of the measured denitrification activity. However, there are also sites where the abundance of denitrifying foraminifers is too low to contribute significantly to nitrogen loss via denitrification. In the Sagami Bay (Japan), the small population of foraminifers apparently contributes only 4% to benthic N₂ production (9).

Foraminiferal denitrification challenges measurements of marine denitrification. Methods that rely on, for example, ¹⁵N tracer additions (e.g., the method used in refs. 27, 29, and 33) or on modeling of nitrate pore water profiles may underestimate denitrification of intracellular NO₃⁻ pools and will consequently underestimate true denitrification (1). Methods that rely on incubations, during which nitrate decreases or the N₂/Ar ratio increases in the overlying waters (e.g., the methods used in ref. 31), on the other hand, may capture foraminiferal denitrification (1). Significant foraminifer-mediated denitrification may therefore call for revisions of current estimates of nitrogen loss from the marine environment and the methodologies used to quantify this loss.

In the past 10 years, our understanding of nitrogen-cycling processes and the microorganisms that mediate these processes has advanced significantly. The microbiology of this cycle has been significantly revised with the recognition that key processes are more broadly distributed among the primary domains of life than previously thought (25, 34). Until recently, bacterial nitrification coupled with denitrification was considered the only process directing fixed nitrogen back to the atmosphere as N₂. Today it seems that, in the ocean, anaerobic ammonium oxidation (anammox) could be as important as bacterial denitrification for N₂ formation (35, 36). We can now supplement the enormous phylogenetic and metabolic biodiversity that is hidden in the microbial world with our identification of phylogenetically and geographically widespread nitrate-storing and denitrifying foraminifers and gromiids.

Materials and Methods

Sites Description and Sampling.

Live specimens of benthic foraminifers were collected during different cruises in 2006–2008 from marginal marine environments and open-sea localities (Table S1). Sediment samples were taken either by hand with a scraper at the shallow sites or by multicoring at deeper sites. The top 10 cm of the sediment was collected and immediately sieved (fractions 63 and 150 μm) using water at in situ temperature and oxygen concentration. The different fractions were stored at in situ temperature.

Specimen Documentation.

Specimens were identified using a stereomicroscope (Leica MZ 12.5 or Wild Heerbrugg M3) or, when necessary, air dried, coated with gold and examined in a LEO 1450VP scanning electron microscope. Species were identified by reference to the taxonomical literature (Table S2).

Isolation of Living Foraminifers and Gromiids.

Individual live specimens, selected for nitrate analysis, were cleaned with a brush in low-oxygen, nitrate-free artificial seawater, transferred to sealed PCR tubes, and analyzed immediately or stored at –20°C until analysis. Denitrification measurements were carried out immediately on live specimens harvested directly from the sediment. Viability of each specimen was assessed by color, the amount of cell cytoplasm, and the gathering of organic matter around the aperture or pseudopodial movement.

Measurement of Nitrate Content.

The intracellular nitrate content of individual species was analyzed using the VCl₃ reduction method (37) as described previously (1, 2) with the following modification: Nitrate was extracted from individual specimens with 10 μl of NaOH (10%) solution in the PCR tubes and was measured in 10-μl subsamples. This procedure avoids disruption of the foraminiferal test, so that shell volume and species identity could be determined after nitrate measurements. Empty foraminiferal tests and sediment corresponding to the volume of the foraminifers were used as contamination controls.

Denitrification Capacity.

Foraminiferal cells were placed in 300 nl chambers with 5 mM Hebes buffer media and nitrate respiration rates were determined from N₂O profiles after acetylene inhibition of N₂O reduction (1, 2, 38, 39). Oxygen in the headspace above the measuring chamber was trapped in an alkaline 0.1 M ascorbate solution, separated from the headspace by a silicon membrane. Oxygen respiration rates were determined using a Clark-type oxygen microsensor (1, 40).

Molecular Phylogenetic Analysis.

To infer the molecular phylogeny of nitrate-storing foraminifers, complete and partial SSU rDNA sequences were taken from the EMBL/GenBank database (access numbers indicated in Table S3) and aligned with Clustal X (41) with manual corrections. The phylogenetic analyses were performed with PhyML (42) under the HKY+I+Γmodel (43). The main alignment is based on a region of SSU situated at the 3′-end (fragment s14F-sB) and includes published sequences from all available genera (45 allogromiids, 19 miliolids, 14 textulariids, and 42 rotaliids) and Gromiida as an out-group.