Source
Hughes, Rodrigues, Bellwood, Ceccarelli, Hoegh-Guldberg, McCook, Moltschaniwskyj, Pratchett, Steneck & Willis (2007), LIT-711, read in NOTE-558: Results and Discussion, Conclusions, Experimental Procedures, Figures 1–4. The Supplemental Data were not read.
What was actually shown
The design. Hughes et al. (LIT-711) worked in Pioneer Bay, Orpheus Island, on the inner Great Barrier Reef, a no-fishing area since 1987, where the 1998 bleaching had sharply cut coral cover and removed branching Acropora from the shallow crest. From 2000 they ran three treatments, four 25 m² replicates each: roofless 4 m tall cages that excluded all large and medium fishes, partial cages open on half their perimeter, and open plots. The experiment lasted 30 months.
The interaction was removed and the community shifted. Herbivorous fish biomass in the cages fell to 0.45 kg/m² per hour of video, against 4.29 and 3.12 outside. Macroalgal cover reached 56% ± 21% after 30 months, against means of 4.1% and 1.7%, and algal biomass was 9 to 20 times higher. Sargassum, absent from the crest before, formed 3 m thickets.
Coral recovery failed through two demographic routes. Coral cover in the cages went from 6.0% to 7.7%, while partial cages and open plots reached 19.2% and 20.2%. Recruits per 25 m² were 39, 108 and 118. Of the colonies present at the start, which had survived the bleaching, 24.2% died in the cages against 9.8% and 11.3% outside. In the cages the count of colonies fell 26%; outside it rose 14–16%. Recovery in the controls was “primarily because of recruitment of species that had been locally extirpated by bleaching”: Acropora was 246 of the 1062 recruits.
This could have come out otherwise. Survivors of the bleaching, branching Porites cylindrica, massive Porites and massive faviids, made up more than 80% of starting coral cover. Had recovery been regrowth of survivors, and had survivors been indifferent to algae, excluding fishes would have changed little. Instead both routes, recruitment and survival, were suppressed. The partial-cage control behaved like the open plots, so the cage structure itself was not the cause.
So the claim. Whether this reef crest returned toward coral dominance after mass bleaching depended on an interaction among its members: large fishes grazing macroalgae. It did not depend only on the corals that survived. The coral community’s recovery is a property of the community, not the sum of colonies’ individual persistence.
What this does not say
- It is not shown for herbivory alone. The cages excluded all large and medium fishes, predators included. Herbivory is identified by the algal response and by the collapse of the algae under grazing once the mesh came off. Small fishes increased in the cages, and the authors suggest they may have helped the seaweeds by cleaning them.
- It is not shown at reef scale. The plots are 25 m², drawing larvae and grazers from an intact surrounding reef. A reef that lost its herbivores everywhere would also differ in larval supply, which the experiment held constant.
- It does not show an alternative stable state. When the mesh came off, macroalgal cover fell from 53% to 13% in 12 days and to about zero in 30. The shifted state did not maintain itself, so “phase shift” here means a shift that lasts while the interaction is absent. It does not mean a hysteretic switch. The record’s account of states held by self-reinforcing loops (THEORY-058) needs the second, and this experiment does not supply it.
- It does not show that local management protects against bleaching. Bleaching was a single event before the experiment and was not manipulated. The authors’ management conclusion is an extrapolation.
- How macroalgae suppress corals is inferred. The authors read the shift of cage recruits toward shade-tolerant Fungia and Euphyllia as an effect of shading. The fecundity and tissue-thickness effects are in the unread supplement.
- One site, four replicates, one post-disturbance window.
Connections
- The individuality holdings. Wilson and Barker (LIT-168) place the coral reef as a group that is a living agent but not a Darwinian individual, and they report the dispute over whether it is an organism or an ecosystem of organisms. This account says the reef’s return to its coral state is maintained by interactions among its members. Both readings of the reef can accommodate that, so it does not decide between them.
- THEORY-058. A contrast in what a shifted state is, set out above. No relation is declared.
- Spawning (THEORY-129). Recovery through recruitment presupposes larval supply, which on the Great Barrier Reef comes from synchronised mass spawning. The two compose; neither rests on the other.