Contribution
A long review of the coral holobiont (host, Symbiodiniaceae, bacteria, viruses, archaea, protists, endolithic algae and fungi), organised to support one proposal. Coral bleaching should be defined as a succession of three holobiont states, the first two “adaptive”, rather than as the loss of algae alone. It adds the “adaptive dysbiosis hypothesis” (ADH) as a name for the first two states. It adds no data.
Key insight
Bleaching is the visible end of a longer trajectory in a multi-partner system. Before the algae are lost, the microbial community shifts. It may buffer the stress by keeping functions going (nitrogen fixation, antioxidants, pathogen exclusion), and it may make room for a different algal partner. Bleaching is either a transient phase of reorganisation or the point where every buffer has failed. Whether it is one or the other depends on the stress’s strength and duration.
Assumptions
- Eubiosis and dysbiosis. Eubiosis is the “healthy state of the holobiont”, dysbiosis the “unhealthy, disrupted state” (§1). The model then applies “dysbiosis” to stages in which the holobiont “remains healthy” (Figure 3 caption). So in the model the word means a departure from the initial microbiome, not ill health.
- Functional redundancy confers resilience, taken from a review of the human gut (Sommer et al. 2017) and applied to corals by analogy (§3).
- Resilience at two levels. At ecosystem level it is the capacity to absorb stress “and to build an alternative state, such that the system maintains its structure and functions”. At holobiont level it is the host’s capacity to resist or revert cellular damage, plus the capacity to recruit new symbionts and restore microbial function (§3).
- Scope: hermatypic Scleractinia.
Key results
All cited; this review reports them, it does not produce them.
- Bleaching routes (§1). The host immune system activates apoptosis, autophagy, exocytosis, detachment or necrosis and rejects the algae. Alternatively, pigments are lost when thylakoids are exposed to free radicals. ROS production in the algae is “the most likely cause” of expulsion. The mechanism is “still far from being fully understood”.
- Host regulation of symbionts (§2.2). “The coral host has the ability to regulate the density and diversity of their associated Symbiodiniaceae”, through tissue compounds, deterioration or digestion of symbionts, and limits on nitrogen and phosphorus.
- Bacterial regulation of symbionts (§2.3.1). Nitrifying and denitrifying bacteria regulate dinoflagellate density by the nitrogen to phosphorus balance. The review states that anammox “constrains” the algae to use N₂, which as written is chemically doubtful, since dinoflagellates do not fix N₂ (see Corrections). Endozoicomonas metabolises algal DMSP.
- Microbial change during bleaching (§2.3.1). Diversity rises, Endozoicomonas falls and Vibrio rises, with virulence genes. Vibrio went from about 0 to 30% relative abundance in heat-bleached corals and back to about 0 in resilient ones (Bourne et al. 2008). Cause or consequence: “impossible to disentangle”.
- Buffers (§§3.1–3.2). Cyanobacteria and plastid ratios increase under stress. Diazotrophs proliferate at high temperature, and their abundance correlates with dinoflagellate abundance. Heterotrophic feeding meets 0–35% of needs in Porites and up to 100% in Montipora capitata during bleaching, and is not enough under recurrent bleaching. Durusdinium confers heat tolerance at a cost in growth, and corals revert to their original partners 2–3 years after a bleaching event.
- Death and after (§3.3). Under abrupt or prolonged heat the microbiome can no longer resist opportunists. After mortality, nitrogen production from dead corals can be as much as 30 times higher, which favours algae, and endoliths speed dissolution of the skeleton.
- Manipulation (§4). A consortium of putatively beneficial bacteria (five Pseudoalteromonas strains, Halomonas taeanensis and a Cobetia marina-related strain) lessened bleaching effects in Pocillopora damicornis (Rosado et al. 2019).
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Bleaching is the breakdown of the obligate coral–dinoflagellate symbiosis, reached by several host or algal routes, with mechanism not fully understood | moderate (consensus as reviewed) | §1, its references 38–42 |
| C2 | Host and bacteria both regulate symbiont density and nutrient exchange | moderate (cited mechanisms, mostly reviews) | §§2.2, 2.3.1 |
| C3 | Bleaching is accompanied by bacterial community shifts (Endozoicomonas down, Vibrio up) | moderate as correlation | §2.3.1, Bourne et al. 2008 |
| C4 | Those shifts cause or buffer bleaching | weak: the authors say causation cannot be determined | §2.3.1, §3.1 |
| C5 | Bleaching is best defined as three successive dysbiosis stages, two adaptive and one traumatic | proposal, untested; no criteria for the stages | §3, Figure 3 |
| C6 | Durusdinium acts as a transitional rather than permanent partner | weak to moderate (one cited reversion study) | §3.2 |
| C7 | The coral holobiont “can therefore be considered as a selective unit” | weak: one citation on transmission of stress-induced microbiota changes, and the units-of-selection roles are not distinguished | §3.1 |
| C8 | Traumatic dysbiosis at scale leads to algal phase shifts on reefs | asserted (Figure 3 caption), not argued in the text | Figure 3 |
Concepts
- eubiosis / dysbiosis — healthy versus disrupted holobiont state. The model also uses dysbiosis for any departure from the initial microbiome.
- adaptive dysbiosis hypothesis (ADH) — microbial restructuring under stress as a mechanism of environmental adaptation, adaptive “as long as the coral host can quickly revert back to a healthy state”.
- metagenomic plasticity — gain or loss of microbial members and genes, or changes in their abundance, within a host generation.
- shuffling / switching — changing the proportions of resident Symbiodiniaceae, versus acquiring new ones from the environment.
- traumatic dysbiosis — irreversible symbiont loss with invasion by opportunists, ending in death.
Connections
- Goulet et al. (LIT-709). Both treat the coral as a consortium. Goulet et al. warn that the Adaptive Bleaching Hypothesis became “coral-centric” through wording. This review takes the hypothesis as the basis of stage 2 and writes it in host-centred terms. Goulet et al.’s octocoral case, where the same symbiont-density drop came by different routes, supports this review’s C1 and is not cited by it.
- Hughes et al. (LIT-711). The Figure 3 caption ends in reef phase shifts to algae. Hughes et al. is experimental evidence for how that happens at reef scale after bleaching: through the loss of herbivory, not through anything in the coral’s microbiome. The review cites other Hughes papers (2003, 2010, 2017, 2018), not this one.
- Lloyd (LIT-160). C7 conflates roles. Transmission of microbiota is a reproducer question, and the review offers no interactor-level evidence.
- Baedke (LIT-167, NOTE-148). The stage-2 reorganisation resembles his “overcomer” (destabilising and reorganising one’s own organisation), carried out by a holobiont. The review does not claim it is controlled by the holobiont, and the resemblance is the record’s observation.
- Borsboom (LIT-547) and THEORY-058. The vocabulary of eubiosis versus dysbiosis as states of a network of interacting parts parallels the network account of disorder as an alternative stable state. The review claims no hysteresis, and stage 2 is explicitly reversible. No relation.
Bearing on the record
- Source, with Goulet et al., for the theory that coral bleaching is the breakdown of a regulated host–symbiont exchange reached by more than one route (THEORY-130). That theory takes C1 and C2 and declines C4 and C5, the microbial-driver claim and the three-stage definition.
- No ML instruction; nothing for the anthology. The probiotic and microbiome-engineering suggestions of §4 are reef management.
Limitations
- No new data. A narrative review; no inclusion criteria for the literature.
- Microbial causation is conceded to be unestablished, yet the title defines bleaching by it.
- The stages are not operational. Nothing says how to tell stage 1 from stage 2 in a sampled coral, or how to measure “buffering capacity”.
- “Adaptive” is asserted, not tested. No fitness comparison is cited between corals whose microbiome shifted and those whose did not.
- The analogy to the human gut (functional redundancy) is imported, not shown for corals.
Open questions
- Does manipulating the microbiota before heat stress change whether, when or how far a coral bleaches? Rosado et al. 2019 is the start. A replicated experiment with microbiome-only manipulation and bleaching as outcome would test C4.
- Can the three stages be distinguished by a measurable signature, so that the model makes predictions?
- Who controls stage-2 reorganisation: host selection, symbiont competition or environmental availability?
Corrections
- Nitrogen chemistry, §2.3.1. The review says that anammox “constrains the dinoflagellate algae to use dinitrogen (N₂) for photosynthesis”. Symbiodiniaceae are not known to fix N₂; diazotrophic bacteria do, as the next sentences of the same section say. The passage probably means that the algae are limited by nitrogen when bacteria remove ammonium. The record does not repeat the sentence as stated.
- “Madaptive” in the §3.3 heading is a typo for “maladaptive”.