Since neutral loci do not show structure, it is likely that propagules disperse extensively around the Atoll. Yet only a subset of them, which carries outlier loci, may survive in specific conditions. Most shifted reefs are within the lagoon and in front of villages. It may be the case that water conditions (high temperatures and nutrients, limited mixing) are the limiting factors that select specific genotypes, allowing Montipora sp. to thrive, while weakening other species of corals, thus shifting the balance among species. Importantly, the type of reef (pristine vs phase-shifted) and localities are correlated, because most phase-shifted reefs happen to be in the north of the Atoll. It is therefore impossible to definitively tease out the potential effects of selection and locality.

The fact that only two individuals out of 138 were genetically identical suggests that sexual reproduction is likely to be a dominant feature of Montipora sp.1 aff. capitata in Ulithi. This is counter to the common local belief that Montipora propagated quickly among Ulithi locations due to human movement between islands and unintentional fragmentation due to anchoring activities.

This may be due to some low level of introgression via gene flow between species, or some remnants of unsorted loci. The one sample from Fiji does contain some genetic material from the blue cluster (Fig. 2) and is unlikely to have experienced recent gene flow with Ulithi individuals. It is therefore more likely that the small amount of genetic material from alternative genetic clusters, as seen in a few individuals, is the result of unsorted ancestral shared loci.

We found that two individuals collected next to each other were genetically identical (population Yealil-Inside, samples 12 and 13), we therefore removed sample 13 and kept sample 12. All other individuals were genetically unique and were included in subsequent analyses.

In order to ensure that clones (individuals resulting from asexual reproduction) were not sampled multiple times, kinship was assessed using GENODIVE (Meirmans 2020).

Suggested that Montipora sp. first spread at landing sites at the four inhabited islands of the Atoll, leading to a conclusion that the breaking of fragile corals associated with anchoring, and the attachment of fragments to anchor lines may have aided the dispersal of Montipora sp. from one island to the next.

This situation is mirrored by fish diversity and biomass, where diverse reefs and Montipora-dominated reefs harbor high or low diversity and abundance of reef fishes, respectively (Crane et al. 2017).

In response to this ecological downgrading, and the continued decline of coral reefs globally, attention has turned to ‘coral gardening’ and reef restoration, focusing on ‘planting’ resistant corals (Camp et al. 2018; Schmidt-Roach et al. 2020). Our work here describes a ‘strong’ or ‘resistant’ coral that is contributing to a phase shift, with negative ecological consequences, a potential cautionary tale for facilitated restoration.

This study shows that selection and adaptation may be contributing to the success of a stony coral (e.g., Phase shift).

We assembled a de-novo reference genome and used RAD seq data with thousands of SNPs to determine if different reefs result from sexual or asexual reproduction, if weedy Montipora fragments are transported between islands by human activities, and if there is evidence of natural selection on specific genotypes, thus favoring spreading success.