Functional Traits Resolve Mechanisms Governing the Assembly and Distribution of Nitrogen-Cycling Microbial Communities in the Global Ocean

Abstract

Microorganisms drive much of the marine nitrogen (N) cycle, which jointly controls the primary production in the global ocean. However, our understanding of the microbial communities driving the global ocean N cycle remains fragmented. Focusing on "who is doing what, where, and how?", this study draws a clear picture describing the global biogeography of marine N-cycling microbial communities by utilizing the Tara Oceans shotgun metagenomes. The marine N-cycling communities are highly variable taxonomically but relatively even at the functional trait level, showing clear functional redundancy properties. The functional traits and taxonomic groups are shaped by the same set of geo-environmental factors, among which, depth is the major factor impacting marine N-cycling communities, differentiating mesopelagic from epipelagic communities. Latitudinal diversity gradients and distance-decay relationships are observed for taxonomic groups, but rarely or weakly for functional traits. The composition of functional traits is strongly deterministic as revealed by null model analysis, while a higher degree of stochasticity is observed for taxonomic composition. Integrating multiple lines of evidence, in addition to drawing a biogeographic picture of marine N-cycling communities, this study also demonstrated an essential microbial ecological theory-determinism governs the assembly of microbial communities performing essential biogeochemical processes; the environment selects functional traits rather than taxonomic groups; functional redundancy underlies stochastic taxonomic community assembly. IMPORTANCE A critical question in microbial ecology is how the complex microbial communities are formed in natural ecosystems with the existence of thousands different species, thereby performing essential ecosystem functions and maintaining ecosystem stability. Previous studies disentangling the community assembly mechanisms mainly focus on microbial taxa, ignoring the functional traits they carry. By anchoring microbial functional traits and their carrying taxonomic groups involved in nitrogen cycling processes, this study demonstrated an important mechanism associated with the complex microbial community assembly. Evidence shows that the environment selects functional traits rather than taxonomic groups, and functional redundancy underlies stochastic taxonomic community assembly. This study is expected to provide valuable mechanistic insights into the complex microbial community assembly in both natural and artificial ecosystems.

Keywords: community assembly; diversity patterns; functional redundancy; functional traits; marine nitrogen cycle; stochasticity.

FIG 1

The composition of N-cycling functional traits and taxonomic groups in the global ocean as revealed by the Tara Oceans metagenomes. (A) The relative abundance of microbial functional traits in different oceans and layers. Only representative gene families with high relative abundances were annotated in the figure. The exact relative abundance for each gene family can be found in Table S2. (B) The relative abundance of microbial phyla in different oceans and pelagic zones. Here, Proteobacteria was further divided into Alpha-, Beta-, and Gammaproteobacteria. (C) The composition of microbial phyla mediating different N-cycling pathways. Here, the same color code as in panel B was used. (D) The relative abundance of N-cycling pathways in epipelagic and mesopelagic zones. Significant differences between epipelagic and mesopelagic are marked with asterisks (**, P < 0.01; ***, P < 0.001). NPO, North Pacific Ocean; SPO, South Pacific Ocean; NAO, North Atlantic Ocean; SAO, South Atlantic Ocean, SO, Southern Ocean; MS, Mediterranean Sea; RS, Red Sea; IO, Indian Ocean; SRF, surface water layer; DCM, deep chlorophyll maximum layer; MES, mesopelagic zone. DNRN, dissimilatory nitrate reduction to nitrite; DNRA, dissimilatory nitrite reduction to ammonia; ANRN, assimilatory nitrate reduction to nitrite; ANRA, assimilatory nitrite reduction to ammonia.

FIG 2

Biogeographic diversity patterns of N-cycling taxonomic groups and functional traits in the ocean. The vertical and latitudinal diversity patterns of taxonomic richness (AB) and functional trait richness (CD), as well as the community dissimilarity and the distance-decay relationship of taxonomic groups and functional traits (EF) were investigated. The black line in panels B and D represents the latitudinal diversity pattern for N-cycling communities across the whole upper ocean. The observed number of microbial species and gene families involved in N cycling is used here for richness. Statistical significance was indicated with asterisks (**, P < 0.01; ***, P < 0.001).

FIG 3

The importance of geo-environmental factors explaining the variations of N-cycling community diversity and composition. (A) Associations between geo-environmental factors and community diversity (Shannon-Wiener index). (B) Associations between geo-environmental factors and community composition (first axis of principal-component analysis). (C) Relationship between different geo-environmental factors (upper right, Pearson correlation coefficient; lower left: Spearman’s rho). (D) Variations explainable by different geo-environmental factors at functional trait level; in the left panel, the associations between the relative abundance of individual functional traits and geo-environmental factors are indicated by the heatmap, whereas the importance of geo-environmental factors in explaining the variations of individual functional traits by random forest analysis is indicated by different sizes of circles; in the right panel, variations explained by the best geo-environmental factor (the one with the largest circle) are indicated by bar plots. DNRN, dissimilatory nitrate reduction to nitrite; DNRA, dissimilatory nitrite reduction to ammonia; ANRN, assimilatory nitrate reduction to nitrite; ANRA, assimilatory nitrite reduction to ammonia. For panels A to C, significance levels for association analyses are marked with asterisks (*, P < 0.05; **, P < 0.01; ***, P < 0.001). For panels A to D, the same scaling color bar was used.

FIG 4

Mechanisms governing the assembly of N-cycling microbial communities. (A and B) Variation partitioning analysis of the contributions of geographic distance and environmental factors in explaining the variations of N-cycling taxonomic groups and functional traits. (C and D) Stochastic ratio representing the stochasticity of community assembly for N-cycling taxonomic groups and functional traits, as revealed by null model analyses. At both the global scale and individual layers, higher stochasticity could be observed for the community assembly of taxonomic groups than functional traits.

FIG 5

Conceptual models illustrating the community assembly and distribution of N-cycling pathways in the oceanic ecosystem. (A) A conceptual model for N-cycling community assembly. First, a regional species pool is formed, adapting to ecological niches in the ocean. Second, the ecosystem selects functional traits rather than species, unless they are highly specialized. Third, functional redundancy of microbial species leads to stochastic community assembly. In the model, different shapes represent different ecosystem functions, whereas different colors represent different microbial taxa. (B) A schematic model illustrating the distribution of N-cycling pathways in ocean. N-cycling pathways differed vertically by depth, instead of by ocean. Relative abundances of functional traits involved in N₂ fixation, organic decomposition, and ANRA were significantly enriched in epipelagic zones, whereas those involved in nitrification, DSRN, and annamox were significantly enriched in the mesopelagic zone. Denitrification is highly detected in both epipelagic and mesopelagic zones but dominated by different functional traits. Specifically, nirS was significantly more abundant in epipelagic zones, while nirK was more abundant in the mesopelagic layer. Ammonia oxidation (NH₄⁺ → NH₂OH) was mainly carried out by archaea.