Genetic patterns in Montipora capitata across an environmental mosaic in Kāne'ohe Bay, O'ahu, Hawai'i

Abstract

Spatial genetic structure (SGS) is important to a population's ability to adapt to environmental change. For species that reproduce both sexually and asexually, the relative contribution of each reproductive mode has important ecological and evolutionary implications because asexual reproduction can have a strong effect on SGS. Reef-building corals reproduce sexually, but many species also propagate asexually under certain conditions. To understand SGS and the relative importance of reproductive mode across environmental gradients, we evaluated genetic relatedness in almost 600 colonies of Montipora capitata across 30 environmentally characterized sites in Kāne'ohe Bay, O'ahu, Hawaii, using low-depth restriction digest-associated sequencing. Clonal colonies were relatively rare overall but influenced SGS. Clones were located significantly closer to one another spatially than average colonies and were more frequent on sites where wave energy was relatively high, suggesting a strong role of mechanical breakage in their formation. Excluding clones, we found no evidence of isolation by distance within sites or across the bay. Several environmental characteristics were significant predictors of the underlying genetic variation (including degree heating weeks, time spent above 30°C, depth, sedimentation rate and wave height); however, they only explained 5% of this genetic variation. Our results show that asexual fragmentation contributes to the ecology of branching corals at local scales and that genetic diversity is maintained despite strong environmental gradients in a highly impacted ecosystem, suggesting potential for broad adaptation or acclimatization in this population.

Keywords: Montipora capitata; Kāne'ohe Bay; clonality; environmental mosaic; genetic relatedness; seascape genomics.

FIGURE 1

Study system. (a) Map of Kāne'ohe Bay with colours denoting 30 sites distributed across five blocks. Sites were chosen using a random stratified design on patch reefs and ~20 Montipora capitata colonies were sampled from each. (b) Example of an individual site, where corals are selected along a 10‐m transect running in up to four cardinal directions from the centre. Photomosaic background for example only, corresponding to a 6‐m radius from the central cinder block.

FIGURE 2

Environmental characteristics. (a) Temperature profiles from August 2017 to August 2019 for the 30 sites, colour coded by block. (b) PCA of all temperature data in panel (a), colour coded by block. Ellipses are 95% confidence intervals. (c) Minimum and maximum hourly temperature at all sites. Grey bar width and point size correspond to nominal site depth and colour corresponds to degree heating weeks (DHW) accumulated over three summers (2017–2019). (d) Mean sedimentation from seven time points on square root transformed y‐axis for visualization. (e) Mean wave velocity in each block. Boxplots represent mean ± 1 IQR.

FIGURE 3

Patterns of clonality in Montipora capitata across Kāne'ohe Bay. (a) Distribution of pairwise genetic distances in biological replicates (69). Red highlighted values were visually inspected and designated as outliers. Grey values represent distribution from which the 95th percentile was calculated to determine clones in the broader population. (b) Identity by descent dendrogram calculated using complete hierarchical clustering. Orange groupings represent biological replicates (n = 10) and blue groupings represent inferred clones. One genotype at site 5_6 was composed of eight colonies, but most other clones were pairs of colonies within the same site.

FIGURE 4

Distance‐based redundancy analysis (dnRDA) was used to examine environmental drivers of underlying genetic patterns. Points represent individual colonies in the study, with only one random individual from each clonal group (i.e., clones removed). (a) Significant (Bonferroni p < .05) environmental drivers of genetic patterns in the bay. Each arrow signifies the multiple partial correlation of the environmental driver in the RDA whose length and direction can be interpreted as indicative of its contribution to the explained variation. Environmental variables are listed in descending order of variance explained. (b) Distribution of colonies based on genotype likelihoods, coloured per block. Ellipses are 95% confidence intervals. (c) Distribution of corals with clonal replicates in the bay. For all dbRDA, only one individual sample from a genotype was used to avoid clonality effects in the analysis.

FIGURE 5

Spatial distribution of genetic and environmental outcomes. (a) Map of Kāne'ohe Bay with colours denoting five blocks and circle size displaying depth. (b) Genet: ramet ratio at each site. (c) Mean pairwise relatedness at each site, excluding all but one representative of each genotype. (d) Mean wave height from a single site in each block. (e) Hours over 30°C at each site. (f) Average residual temperature at each site.