The ocean’s ammonia oxidizers are not distributed evenly—and the best-known habitats may be giving us a misleading view. This study finds a larger, more diverse archaeal world, with much of it still unnamed. Ammonia-oxidizing archaea, or AOA, and ammonia-oxidizing bacteria, or AOB, are abundant in the global ocean and play a crucial role in marine nitrogen cycling through nitrification.

Despite extensive studies using the ammonia monooxygenase alpha-subunit gene, called amoA, understanding of their distribution patterns in marine habitats remains limited. The study provides an integrative framework for assessing diversity and biogeography across marine habitats through a meta-analysis and phylogenetic examination of archaeal and bacterial amoA sequences from public databases.

Most sequences came from estuarine sediments and coastal waters, while the pelagic open ocean, oxygen minimum zones, deep sea, hydrothermal vents, and benthic organisms remain underexplored. Aerobic ammonia oxidation to nitrite is the critical first step in nitrification and plays a vital role in the nitrogen cycle.

In marine ecosystems, ammonia primarily originates from organic matter decomposition, marine organism excretion, and anthropogenic sources. Once ammonia is oxidized to nitrite, nitrite serves as an energy source for the second step of nitrification and for processes such as anammox and denitrification.

Ammonia-oxidizing microorganisms therefore help regulate nitrogen substrates and marine nitrous oxide production. Ammonia oxidation is primarily mediated by two groups of aerobic autotrophic microorganisms: ammonia-oxidizing archaea and ammonia-oxidizing bacteria.

AOA comprise a diverse group within Nitrososphaeria of Thaumarchaeota, classified into four basal lineages that include genera such as Nitrosopumilus, Nitrosoarchaeum, and Ca. Nitrosopelagicus. AOB include Nitrosomonas and Nitrosospira among beta-proteobacteria, and Nitrosococcus among gamma-proteobacteria.

This study focuses on canonical ammonia oxidizers because reports of comammox Nitrospira in marine environments remain limited and primarily restricted to coastal habitats. Ammonia monooxygenase catalyzes the oxidation of ammonia to hydroxylamine.

This membrane-bound enzyme is encoded by the amoABC operon, with amoA being the most widely used marker gene for studying ammonia oxidizers. Research during the last two decades using amoA has clarified ecological characteristics of AOA and AOB and their contributions to marine nitrogen cycling.

Historically, AOB were considered the primary drivers of ammonia oxidation in the ocean, but the discovery of AOA challenged that paradigm. Studies show that AOA are generally more abundant and widespread in marine environments, particularly oligotrophic waters, whereas AOB thrive in high-ammonia environments such as eutrophic coastal areas.

The competitive advantage of AOA in nutrient-poor environments is attributed to their high ammonia affinity. However, the pattern is not clear-cut, because some AOA, including Ca. Nitrosocosmicus, are adapted to high ammonia concentrations. The distribution of AOA and AOB in marine ecosystems is shaped by environmental factors, including oxygen availability.

AOA generally exhibit lower oxygen requirements than AOB. As a result, AOA are also dominant in oxygen-depleted zones such as oxygen minimum zones, where their activity might be coupled to anammox or denitrification. The complex interplay of environmental factors creates diverse ecological niches for marine ammonia-oxidizing communities.

Investigating their diversity and distribution can clarify the mechanisms regulating their roles in marine nitrogen cycling and help predict responses to environmental changes. Current knowledge of their distribution patterns remains relatively limited.

The study addresses this gap with a comprehensive overview of AOA and AOB diversity and distribution in marine ecosystems. It does so by compiling a curated reference database, conducting a meta-analysis, and performing a phylogenetic analysis of amoA sequences across marine environments.

Archaeal and bacterial amoA sequences were extracted from articles and unpublished references obtained from Google Scholar, Scopus, and GenBank through October 2021. The search combined amoA, archaea or bacteria, and terms related to marine habitats such as estuary, coast, open ocean, deep sea, coral, and hydrothermal vent.

Marine sequences were selected from molecular techniques based on Sanger sequencing, including clone libraries, T-RFLPs, and DGGE. The database contained 11,411 archaeal amoA sequences from 49 articles and 13 unpublished references, representing 81 marine sites. It also contained 7,031 bacterial amoA sequences from 46 articles and 2 unpublished references, corresponding to 59 marine sites.

The sequences were classified by marine environment and habitat, including water column, sediment, benthic organisms, ocean ridge, estuary, coast, pelagic open ocean, oxygen minimum zone, deep sea, coral, sponge, hydrothermal vent, and seawater aquarium. A map of study sites for each archaeal and bacterial amoA gene was constructed using Ocean Data View five point three point zero.

The upper AOA map shows especially dense clustering of sites in the North Atlantic, Mediterranean, and western Pacific, with additional coverage across the Southern Ocean. The lower AOB map is more concentrated along continental margins and semi-enclosed seas, while the central Pacific and Southern Ocean contain relatively few symbols; both panels use the same water, sediment, benthic-organism, and ocean-ridge markers for comparison.

Archaeal and bacterial amoA sequences were processed separately with QIIME two version 2021 point 11. Using the de novo clustering plugin with the VSEARCH algorithm, sequences were dereplicated and clustered into operational taxonomic units at a 95 percent similarity threshold.

De novo chimera detection was checked with VSEARCH. For AOA and AOB separately, the ClustalW algorithm in BioEdit version 7.2.5 aligned representative operational taxonomic units with published reference amoA sequences from the NCBI database. A maximum likelihood tree for AOA and another for AOB was constructed in MEGA-XI using the Tamura-Nei model and 10,000 bootstrap replicates.

Venn diagrams and UpSet diagrams were constructed in R for AOA and AOB to identify the number of operational taxonomic units found in each environment and habitat and their degree of overlap. Non-metric multidimensional scaling analyses were also performed using a Jaccard dissimilarity matrix with the vegan package.

The AOA studies were conducted mainly in the water column and sediments, with fewer studies of benthic organisms and ocean ridges. The meta-analysis found 641 AOA operational taxonomic units among 10,823 good-quality sequences from 81 marine sites. AOA sequences came mainly from sediments, followed by water, benthic organisms, and ocean ridges.

For AOB, the meta-analysis found 224 operational taxonomic units among 6,738 good-quality sequences from 59 marine sites. Most AOB sequences came from sediments, followed by water, benthic organisms, and ocean ridges. The analysis identified 25 phylogenetic clades of marine AOA, and 52 percent of sequences, representing 386 operational taxonomic units, were unassigned to known lineages.

Among clades affiliated with cultured AOA, the most abundant were Ca. Nitrosopelagicus, Nitrosoarchaeum, and Nitrosopumilus. Ca. Nitrosopelagicus included 52 operational taxonomic units and 18.6 percent of sequences; Nitrosoarchaeum included 65 operational taxonomic units and 13.9 percent; and Nitrosopumilus included 63 operational taxonomic units and 9.5 percent.

In the circular tree, the water-associated dots concentrate around clusters I, II, IV, and XXV, while sediment-colored dots are especially widespread through the central and lower branches. The habitat pie charts reinforce that contrast: cluster XXV is dominated by estuary and coastal samples, whereas several Nitrosoarchaeum-associated branches show strong estuary representation alongside other marine habitats.

Most AOA sequences came from estuaries and coasts, followed by the deep sea, pelagic open ocean, and oxygen minimum zones. Fewer sequences came from corals, sponges, marine aquariums, and hydrothermal vents. Ca.

Nitrosopelagicus was widespread across marine habitats except aquariums, while Nitrosopumilus was more common in aquariums. Nitrosoarchaeum had high relative abundance in coastal areas, whereas Ca. Nitrosotalea, Nitrososphaera, Ca.

Nitrosocosmicus, and Ca. Nitrosotenuis were highly abundant in the deep sea. In the upper archaeal panel, the estuary-colored segments occupy substantial portions of several genera, while the bacterial panel shows only four genera and a more concentrated distribution of habitat colors.

The sequence totals printed beneath the bars also differ markedly: the archaeal counts range from seventy-three to two thousand eighteen, whereas the bacterial counts range from five to five thousand seventy-three. A large proportion of AOA-amoA sequences remained unclassified across habitats.

Only 23.9 percent of hydrothermal vent sequences and 25.4 percent of deep-sea sequences could be assigned to known genera. Only 28.3 percent of sponge sequences were classified, while 51 percent of coral sequences were attributed primarily to Ca. Nitrosopelagicus and Ca.

Nitrosocaldus. Aquarium samples had the highest proportion of classified sequences, dominated by Nitrosopumilus at 70.3 percent. In the AOB panel, the estuary pie is divided between Nitrosomonas and unclassified sequences, whereas the coast and deep-sea pies are overwhelmingly unclassified.

The smaller habitat samples are even more concentrated: the OMZ and hydrothermal-vent charts are entirely unclassified, while the sponge chart includes only narrow classified slices. The analysis identified 13 phylogenetic clades of marine AOB. Clades II and IV were affiliated with Nitrosospira, clades VII, X, XII, and XIII with Nitrosomonas, and clade VI with Nitrosococcus.

A small clade containing three operational taxonomic units and five sequences remained unclassified. Nitrosospira was the dominant AOB, with 107 operational taxonomic units and 75.3 percent of sequences, followed by Nitrosomonas and Nitrosococcus.

The pie charts around the tree show that clade five contains the largest sequence total, with a predominantly coastal-colored distribution, while the smaller clades have visibly more mixed habitat composition. The branch colors and Roman numerals also separate the Nitrosospira-associated groups in the upper arc from the Nitrosomonas-associated groups spanning the lower and left portions.

AOB sequences were retrieved mainly from estuaries and coasts. Fewer AOB sequences were obtained from the deep sea, sponges, hydrothermal vents, and the pelagic open ocean. AOA sequences were classified into 641 operational taxonomic units, with the highest number found in sediments: 394 operational taxonomic units, 72.8 percent unique.

The water column contained 263 operational taxonomic units, benthic organisms contained 88, and ocean ridges contained 51. Sediments shared 82 operational taxonomic units with water, 31 with ocean ridges, and 28 with benthic organisms. In the AOA UpSet plot, the tallest bars at left correspond to estuary and deep-sea habitat counts, while the many short bars across the right show numerous smaller habitat combinations.

In the AOB panel, the overlap structure is more concentrated: the estuary and coast rows dominate the bottom counts, and most intersection bars remain comparatively low. The NMDS analysis showed that the AOA community from Northern Baffin Bay in the Arctic Ocean was distinct from all other communities.

The remaining communities were structured by two marine environments, with sediment-associated AOA forming a more dispersed cluster than water-associated AOA. Within sediments, communities separated into deep-sea sediments and sediments from estuarine and coastal regions.

These clustering patterns by environment and habitat were confirmed by PERMANOVA and ANOSIM, with a p value of zero point zero zero one. However, low R-squared values from PERMANOVA indicate that a substantial portion of variation remains unexplained, suggesting that additional factors may influence AOA distribution.

Environment, habitat, and climate zone were significant factors influencing marine AOB distribution, with a p value of zero point zero zero one. However, low R-squared values from PERMANOVA indicate that additional factors may contribute to AOB distribution.

Marine AOA exhibited higher phylogenetic diversity than AOB, with 25 clades compared to 13 clades. The disparity can be partially attributed to the greater number of recovered AOA-amoA sequences, while also aligning with previous studies indicating that AOA are generally more abundant than AOB in oceanic environments.

These findings suggest that AOA possess higher genetic and ecological diversity, allowing them to thrive in a wide range of marine conditions. Seventeen phylogenetic groups of AOA, comprising over half of the total amoA sequences, remained unclassified, with particularly high proportions in ocean ridges and sponges.

The findings show that a significant portion of marine AOA diversity remains unexplored and highlight the need for further isolation and characterization, particularly from sponges, hydrothermal vents, and the deep sea. Most amoA sequences were recovered from estuarine sediments and coastal waters, while benthic organisms, ocean ridges, the pelagic open ocean, and deep-sea habitats remain largely unsampled.

Marine AOA exhibited higher phylogenetic diversity than AOB, with many unclassified AOA-amoA sequences that may represent novel AOA clades. The widespread occurrence of AOA, particularly Ca. Nitrosopelagicus and Nitrosopumilus, underscores their adaptability to oligotrophic and extreme environments such as oxygen minimum zones and deep-sea regions.

Conversely, marine AOB, primarily Nitrosospira and Nitrosomonas, are more prevalent in nutrient-rich environments such as estuarine and coastal sediments. Further research using advanced sequencing and multi-omics approaches is needed to investigate AOA and AOB in under-sampled marine habitats.

AOA occupy a broader range of marine conditions and show greater phylogenetic diversity than AOB, while major sampling gaps leave deep, pelagic, vent, and benthic habitats poorly understood.