Multi-species consumer jams and the fall of guarded corals to crown-of-thorns seastar outbreaks

Outbreaks of predatory crown-of-thorns seastars (COTS) can devastate coral reef ecosystems, yet some corals possess mutualistic guardian crabs that defend against COTS attacks. However, guarded corals do not always survive COTS outbreaks, with the ecological mechanisms sealing the fate of these corals during COTS infestations remaining unknown. In August 2008 in Moorea (17.539° S, 149.830° W), French Polynesia, an unusually dense multi-species aggregation of predators was observed feeding upon guarded corals following widespread coral decline due to COTS predation. Concurrent assaults from these amplified, mixed-species predator guilds likely overwhelm mutualistic crab defense, ultimately leading to the fall of guarded corals. Our observations indicate that guarded corals can sustain devastating COTS attacks for an extended duration, but eventually concede to intensifying assaults from diverse predators that aggregate in high numbers as alternative prey decays. The fall of guarded corals is therefore suggested to be ultimately driven by an indirect trophic cascade that leads to amplified attacks from diverse starving predators following prey decline, rather than COTS assaults alone.

Identifying ecological processes that drive species trajectories is a prerequisite for ecosystem management. However, community dynamics are sometimes governed by unexpected, indirect interactions and complex emergent properties that can cause runaway responses and abrupt ecological shifts (Silliman et al., 2013;Terborgh & Estes, 2010). Outbreaks of the coral predator crown-of-thorns seastar (COTS) cause widespread coral mortality across the Indo-Pacific Ocean (Pratchett et al., 2014) with often drastic impacts on diverse reef communities (Kayal et al., 2012). However, some coral species possess mutualistic allies that can deter COTS predation. In particular, trapeziid crabs and alpheid shrimps inhabiting large pocilloporids are known for their ability to effectively defend their host corals from COTS assaults   Figure 1). Despite increasing understanding of factors determining coral susceptibility to COTS predation (Glynn, 1976;Kayal et al., 2011;Kayal & Kayal, 2017;Pratchett, 2001;Rouzé et al., 2014), the processes sealing the fate of guarded corals during outbreaks have remained unknown. Here we provide insights into the ecological mechanisms underlying the fall of guarded corals during predatory COTS outbreaks.

Methods
Our observations were performed at the peak of an intense crown-of-thorns seastar (COTS) outbreak that decimated coral communities around the island of Moorea (

Amendments from Version 1
We have updated our manuscript following judicious comments provided by the referees. The new version notably includes further information, as well as an extended discussion, on the reported ecological process and its prevalence in our study system. The new version also provides more details on coral mutualist crustaceans and clearer Figure 1, Figure 2, and Supplementary Image 1. Following comments from the referees, we have added arrows to the figures, in order to clarify which parts we are referring to. We are thankful to the referees for their constructive remarks.

Results and discussion
In August 2008 at 12 m depth on Tiahura reef, we observed an unusually dense aggregation of coral-eating butterflyfishes jamming around guarded pocilloporids, the last coral bastions that had yet resisted swarms of the predatory seastar ( Figure 2, Supplementary Image 1). Widespread coral decline had previously wiped out much of resident populations of coral-feeding butterflyfishes (Kayal et al., 2012), pushing starving survivors to aggregate around the guarded corals. While butterflyfishes were increasingly observed to gather around guarded corals as the COTS outbreak progressed around the island, the aggregation of 9 butterflyfishes within a single square-meter (9 fish.m -2 ), as captured in Figure 2, was particularly surprising. Density of the coral-feeding butterflyfish assemblage on this reef location had dropped to the much lower average value of 4.3±0.9 SE fish.200m -2 following the COTS outbreak (surveyed in June 2008, equivalent to 0.02 fish.m -2 ). The observed aggregation thus represented a more than 400-times concentration of the predation pressure exerted by the butterflyfishes, and was targeting a guarded pocilloporid that was already under attack by COTS (Figure 2, Supplementary Image 1).
Guarded pocilloporids in Moorea have shown the ability to resist devastating COTS predation for several years (McKeon & Moore, 2014; Figure 1). However, concurrent assaults from such locally amplified, mixed-species predatory guilds likely overwhelm the ability of trapeziid crabs and other exo-symbionts to defend host pocilloporids, ultimately causing the fall of guarded corals. Indeed, coral occupation by mutualist communities is determined by strict rules of territoriality and competition (Glynn, 2013;Leray et al., 2012), which limits the abundance of inhabiting guardians in host colonies, and therefore their ability to sustain predatory assaults. The relative contribution of the butterflyfishes, as compared to COTS, to the death of guarded corals at this stage remains unclear. Further research is needed to quantitatively evaluate the aptitude of coral mutualists to withstand attacks from mono-versus multi-specific predators at different abundances. This particularly applies when the predator guilds involve specialized coral-feeding species from distant phyla with different feeding modes, such as fishes that sample polyps through repeated rapid bites and seastars that consume large portions of coral tissue over extended amount of time. Coral decline has already been identified as an engine of COTS movements and prey selection during outbreaks (Kayal et al., 2011;Kayal et al., 2012;Silliman et al., 2013). Our observations suggest that further cascading effects include aggregating diverse predators in numbers surpassing mutualistic defenses, eventually leading to the collapse of guarded corals. We therefore advocate the importance of controlling COTS outbreaks at the earliest stages, before trophic cascades could lead to a runaway collapse of coral communities.

Data availability
All data underlying the results are available as part of the article and no additional source data are required.

Competing interests
No competing interests were disclosed. , also should be noted as defending pocilloporid corals from Alpheus lottini COTS attacks. This shrimp guard occurs world-wide on pocilloporid corals.

Grant information
It would also be worth noting the defensive behaviour, if any, of the crustacean guards toward the fish corallivores.
'White feeding scars' are referred to in Fig. 1 and Fig. 2 (supplementary image). These are difficult to make out in the photographs. I suggest adding arrows to make these easier to see. Also, it would be useful to know the approximate diameters of the colonies.

If applicable, is the statistical analysis and its interpretation appropriate? No
Are all the source data underlying the results available to ensure full reproducibility? Yes

Are the conclusions drawn adequately supported by the results? Yes
No competing interests were disclosed.

Competing Interests:
Referee Expertise: Reef coral biology and ecology.
I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.