Contribution
The first transcriptome time course across a coral’s spawning night, combined with light manipulations that stop or shift spawning. Before it, moonlight was known from field correlation and older manipulations (Babcock et al. 1986; Willis et al. 1985) to time mass spawning, but the molecular events of the night were not described. It adds three things: a set of genes whose expression changes only on the spawning night, evidence that added light or darkness at night abolishes both that programme’s normal timing and the spawning, and a proposed GPCR signalling model for gamete release.
Key insight
The cue does not only set a date. On the night itself, light conditions after sunset decide whether and when a mature colony releases gametes, and a week of altered nights is enough to break it. The transcript programme that accompanies release starts early under light and does not start under darkness. So the timing runs through a light-sensitive process on the night itself, not only through an entrainment set months before. The authors contrast this with Willis et al. 1985, which suggested months.
Assumptions
- Species and site. One species, Acropora millepora, from the Heron Island reef flat (23°33′S, 151°54′E), at the Heron Island Research Station, kept dark at night against stray light.
- Maturity. Colonies were checked for pink eggs before use.
- Tank conditions stand in for the reef. Outdoor flow-through tanks with reef-flat water and natural sun and moon. Ambient colonies spawned when the reef did, which the authors take as validation.
- Transcript selection by variance. The 184 “spawning” genes are those in the top 10% of Poisson-normalised variance on the spawning day and not in the top quartile on earlier days or in the August full and new moon samples, with more than 100 reads in some condition. Other cut-offs (top 5–25% versus not top 25–50%) “produced similar results”. The up-regulated and down-regulated sets require a twofold difference against both the previous days and the August samples.
- Library design. 12 libraries (full and new moon, August) on one lane, each sequenced twice as technical duplicates. 24 libraries from the spawning experiment on two lanes. The paper does not say how 16 colonies, three treatments and up to seven time points map onto 24 libraries, so biological replication per time point is not stated.
Key results
- 2011 manipulation (Figure 2A, Methods). Twenty colonies were collected on 9 November 2011. Four were left on the reef flat (treatment F, not analysed further in the text), and 16 went to tanks: ambient (A, N = 6), light (L, N = 5, about 5 µmol quanta m⁻² s⁻¹ PAR from 18:15 to 24:00, then dark) and dark (D, N = 5, shaded 18:15 to sunrise). Sampling was at noon and at moonrise on 10, 12, 14 and 15 November. On the spawning night of 16 November, samples were taken at noon, 18:15, 19:30, 21:00, 22:00, 22:30 and 00:20. Ambient colonies began “setting” at 19:30 and released gametes between 21:30 and 22:30. “No sign of spawning behavior occurred in either the light or dark treatments.” Whether L and D colonies spawned on later nights is not reported.
- Spawning-day genes (Figures 1C, 2B). 184 transcripts. In A they changed (mostly induced) just before and at release. In D they did not change. In L they changed early, around 18:15 and 19:30. GO enrichment (FDR < 0.1): cell cycle, GTPase activity, signal transduction.
- Up and down sets (Figure 2—figure supplement 1, Figure 3B–C). 177 up-regulated (54 also in the 184) and 29 down-regulated (none in the 184). Up: G-protein-coupled processes, signal transduction, respiration. Down: rhythmic processes including circadian clock-related genes. Induced genes include two opn4b (melanopsin)-like homologs, many GPCRs, a synaptotagmin 7 homolog and a Trp-related protein 4. Expression did not correlate with the released gametes’ own transcriptome (mean Pearson r 0.22, and 0.05 for the spawning-day genes), so the signal is from the colony tissue, not from gametes.
- qPCR check (Figure 2—figure supplement 2). Twelve genes, ambient colonies at 22:00 against 19:30: r² = 0.92 with RNA-seq log₂ fold change, p < 0.0001.
- Lunar transcription (Figure 1—figure supplement 1). August, N = 4 colonies at 2 m depth, at 12:00, 18:00 and 24:00 on new and full moon days. Differences between full and new moon were tested against Poisson noise with a Bonferroni correction. Genes higher at full-moon midnight include cry1, cry2 and thyrotroph embryonic factor.
- 2006 light-quality experiment (Figure 2—figure supplement 3). 90 colonies in tanks shaded by day to simulate 3 m depth, unshaded at night. For six hours after sunset (18:00–24:00) they received high (100), medium (50) or low-dim (0.75–1 µmol quanta m⁻² s⁻¹) white light, or blue, green or red light at 1 µmol quanta m⁻² s⁻¹, from 31 October to 15 November 2006. Full moon was 5 November. The reef and the controls spawned on 13 November, eight nights after full moon. Red and ambient colonies spawned at 21:30. Blue, green and all white intensities delayed spawning 6–8 h or to two nights later. One-way ANOVA with Tamhane post hoc, p < 0.01, for the phase shift.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Altering nocturnal light for the week before spawning (added light after sunset, or darkness) prevents A. millepora colonies from spawning on the spawning night | moderate: clear all-or-none outcome, but N = 5–6 per arm, one night, one site, later nights not reported | Results; Methods (2011) |
| C2 | Blue, green and white nocturnal light delay spawning by hours or nights, and red light does not | moderate: 90 colonies, but the result is shown in a supplementary figure, with treatment numbers that do not add up as written (see Limitations) | Figure 2—figure supplement 3 |
| C3 | A set of transcripts changes only on the spawning night, around gamete release, and its timing follows the light treatment | moderate as description; selection by variance ranking, replication per time point unstated | Figures 1C, 2B |
| C4 | Spawning-night up-regulated genes are enriched for GPCR signalling and include melanopsin-like homologs | moderate (GO enrichment, qPCR-validated fold changes for 12 genes) | Figure 3; Figure 2—figure supplement 2 |
| C5 | Gamete release is triggered by a melanopsin-like photoreceptor and/or neuropeptides acting through GPCR cascades | weak: a model from co-expression; the authors say the photopigment role is not shown | Figure 4; Discussion |
| C6 | Coral transcription varies with lunar phase, including cryptochromes | weak to moderate: N = 4, apparently one library per condition with technical duplicates | Figure 1—figure supplement 1 |
| C7 | Disruption by altered night light takes effect within 7 days, faster than the months of entrainment earlier work proposed | moderate, from C1 and C2 | Discussion; author response |
Concepts
- mass spawning — the annual release of gametes by many coral species over a few nights, a few days after a full moon in the spawning season on the Great Barrier Reef (November in both experiments here).
- setting — gamete bundles appearing in the polyp mouths before release.
- light pollution — here, artificial PAR light after sunset.
- phase shift (of spawning) — the delay, in hours or nights, relative to ambient colonies. Unrelated to the ecological “phase shift” of LIT-711.
Connections
- Hughes et al. (LIT-711) share an author (Hoegh-Guldberg) and the Great Barrier Reef. This paper says reproduction is “one of the most important processes for the persistence of reefs”; Hughes et al. show recovery after bleaching running through recruitment. Neither cites the other.
- Collective emotion (LIT-696, THEORY-126). Von Scheve and Ismer’s minimal collective emotion is “synchronous convergence” across individuals produced by a shared event with no mutual awareness. Mass spawning, on this paper’s evidence, is synchrony of that minimal kind in a non-cognitive system. The parallel is the record’s; no relation is claimed.
- Coupled-oscillator synchrony (THEORY-069, LIT-091). The record’s synchronisation material is about units that pull on each other. Nothing here is coupling, which keeps this case apart from that material.
Bearing on the record
- Source of the theory that in A. millepora the nocturnal light regime is necessary for spawning on the expected night, with a spawning-night GPCR programme whose timing follows the light (THEORY-129). That theory takes C1–C4 and C7. It holds C5 as a proposal and C6 as background.
- No ML instruction; nothing for the anthology.
Limitations
- Single species, single site, single spawning night for the transcriptome (2011), and single season for the light-quality test (2006).
- Small N in 2011 (5–6 colonies per arm), and the fate of the L and D colonies after 16 November is not reported. “Prevented” may be “delayed”, which is what the 2006 experiment found.
- Replication of the RNA-seq is unclear. 24 libraries cannot cover every colony at every time point and treatment, and the text does not say how samples were pooled. The spawning-gene list comes from a variance filter, not a test against biological replicates.
- Inconsistent numbers in the 2006 design. “Five colonies were placed into each aquarium”, “three aquaria for each of the seven treatments” and intensity (n = 15) and spectra (n = 10) groups do not reconcile with 90 colonies. A control group is also mentioned. The treatment sizes are therefore uncertain.
- No mechanism test. No inhibitor, knockdown or receptor assay. The melanopsin-like genes’ photopigment function is untested, as the authors say.
- The cue is not isolated. Both added light and darkness abolished spawning, so the experiment shows that the natural night-light pattern is needed, not which feature of it (moonlight intensity, the dark interval after sunset, moonrise time). The 2006 experiment only adds light, and shows its spectrum matters.
- No test of inter-colony coordination.
- Timeline. The text says treatments began “8 days prior to the spawning night”. The methods give collection on 9 November and spawning on 16 November, which is seven days.
Open questions
- Do colonies kept dark or lit spawn later, and is the delay a fixed number of nights (as in 2006) or abolition for the season?
- Which feature of the night is the cue: irradiance after moonrise, the dark interval after sunset, or the timing between them? A design that shifts artificial “moonrise” while keeping the dark interval would separate them.
- Are the melanopsin-like homologs photopigments, and does blocking GPCR signalling block release?
- Does a coral species that spawns on a different night use the same programme with a different trigger? The reviewers proposed this and the digest names it as future work.