Source
Kaniewska, Alon, Karako-Lampert, Hoegh-Guldberg & Levy (2015), LIT-710, read in NOTE-557: Results and discussion, Materials and methods, Figures 1–4 and their supplements (from captions), the decision letter and the author response.
What was actually shown
The cue is necessary, in this setup. Kaniewska et al. (LIT-710) held 16 reproductively mature A. millepora colonies from the Heron Island reef flat in outdoor flow-through tanks under natural sun and moon, from 9 November 2011 to the spawning night of 16 November. Six were left under ambient conditions. Five received about 5 µmol quanta m⁻² s⁻¹ of artificial light from 18:15 to midnight and darkness after. Five were shaded from 18:15 to sunrise. The ambient colonies spawned between 21:30 and 22:30, as the reef did. The other two groups showed “no sign of spawning behavior”. This could have come out otherwise: mature colonies with gametes already formed could have released them on schedule from an entrainment set earlier, which is what the months-long entrainment of earlier work (Willis et al. 1985) would lead one to expect.
Light quality matters, and the effect is quick. In a separate 2006 experiment with 90 colonies, six hours of nocturnal light after sunset in blue, green or white light, at intensities from about 1 to 100 µmol quanta m⁻² s⁻¹, delayed spawning by 6–8 h or by two nights. Red light did not delay it, and those colonies spawned at 21:30 with the controls and the field (one-way ANOVA, Tamhane post hoc, p < 0.01). Both experiments altered the nights for about a week, so whatever the light acts on responds within days.
A transcriptional programme follows the cue. From 12,384 genes the authors selected 184 that were highly variable only on the spawning day. In ambient colonies these changed, mostly upward, just before and during gamete release. In dark colonies they did not change. In lit colonies they changed early, around 18:15 and 19:30. Up-regulated genes on the spawning night (177) were enriched for G-protein-coupled signalling, signal transduction and respiration. They include two melanopsin-like homologs and many GPCRs, and qPCR on 12 genes matched RNA-seq with r² = 0.92. Down-regulated genes (29) were enriched for rhythmic processes. The programme is in colony tissue, not gametes: its correlation with the released gametes’ transcriptome was near zero.
So the claim. In this species at this site, the night-light conditions of the days before and of the spawning night are necessary for release on the expected night. Displacing them displaces or abolishes both the release and a spawning-night signalling programme that is timed with it. The authors’ proposed cascade runs from moonlight or another signal, through melanopsin-like photoreceptors and neuropeptides acting on GPCRs, to follicle rupture and gamete release (their Figure 4). The account holds it as a hypothesis the expression data are consistent with, not as shown.
What this does not say
- It does not show that moonlight as such is the cue. Both added light after sunset and darkness all night abolished spawning. What the experiment shows is that the natural pattern is needed. It does not show which feature: the dark interval after sunset, moonrise, or moonlight intensity. The authors’ own list of candidate cues includes “the onset of darkness”.
- It does not show the cascade causes release. Nothing was blocked, knocked down or applied. The authors say they have “no direct evidence” that the melanopsin-like homologs are photopigments.
- It does not show an endogenous clock. No colony was held in constant conditions to show a free-running rhythm. Rhythmic-process genes go down on the spawning night and cryptochromes vary with lunar phase (N = 4, August), which is consistent with a clock and does not demonstrate one.
- It does not show coordination between colonies. Each colony responded to the same external drive. Synchrony across the reef, and across the 130-odd species that spawn together on the Great Barrier Reef, is the expected consequence of a shared cue. It is not measured here, and coupling between colonies (for example by a pheromone, which the authors mention once on the strength of a tachykinin-like receptor) is not tested.
- “Prevented” may be “delayed”. The fate of the lit and dark colonies after 16 November is not reported, and the 2006 experiment found delays of up to two nights.
- Scope. One species, one site, one spawning night for the transcriptome, groups of 5–6, and replication of the RNA-seq libraries that the paper does not state.
Connections
- THEORY-126 (collective emotion) has as its minimal case convergence across individuals produced by a shared event, with no mutual awareness and no group subject. On this evidence coral mass spawning is a non-cognitive instance of that minimal form, collective timing by common cause. The parallel is the record’s; neither account is about the other’s phenomenon, and no relation is declared.
- THEORY-069 (HKB) is the record’s account of collective timing by coupling, where units entrain one another and the collective variable undergoes a phase transition. This account is its contrast case: timing by a shared external drive with no coupling shown. No relation is declared.
- Reef recovery (THEORY-128). Recovery after bleaching on the same reef system runs through recruitment, which depends on successful spawning. The two accounts compose; neither rests on the other.