Population structure and clonal prevalence of scleractinian corals (Montipora capitata and Porites compressa) in Kaneohe Bay, Oahu

As the effects of anthropogenic climate change grow, mass coral bleaching events are expected to increase in severity and extent. Much research has focused on the environmental stressors themselves, symbiotic community compositions, and transcriptomics of the coral host. Globally, fine-scale population structure of corals is understudied. This study reports patterns of population structure and clonal prevalence found in Montipora capitata and Porites compressa in Kaneohe Bay, Oahu. Generated using ddRAD methods, genetic data reveals different patterns in each taxa despite them being exposed to the same environmental conditions. STRUCTURE and site-level pairwise FST analyses suggest population structure in M. capitata resembling isolation by distance. Mantel tests show strong, significant FST correlations in M. capitata in relation to geographic distance, water residence time, and salinity and temperature variability (range) at different time scales. STRUCTURE did not reveal strong population structure in P. compressa. FST correlation was found in P. compressa in relation to yearly average sea surface height. We also report high prevalence of clonal colonies in P. compressa in outer bay sites exposed to storms and high energy swells. Amongst only outer bay sites, 7 out of 23 sequenced individuals were clones of other colonies. Amongst all 47 sequenced P. compressa individuals, 8 were clones. Only one clone was detected in M. capitata. Moving forward, it is crucial to consider these preexisting patterns relating to genetic diversity when planning and executing conservation and restoration initiatives. Recognizing that there are differences in population structure and diversity between coral taxa, even on such small-scales, is important as it suggests that small-scale reefs must be managed by species rather than by geography.


44
Rapid climate change due to anthropogenic carbon emissions is one of the greatest threats 45 to global marine biodiversity (Cheung et al. 2009). Within the past few decades, coral bleaching 46 events have increased in occurrence and severity to the point where they are becoming 47 commonplace (Hughes et al. 2003). Despite bleaching being a widely-known impact of climate 48 change, the pathways by which it occurs remain poorly understood. 49 A large proportion of research has focused on the role of zooxanthellae, dinoflagellate 50 algae of the genus Symbiodinium that form symbiotic relationships with coral, in mediating the 51 bleaching response. In a zooxanthellae driven response, thermal bleaching is caused by or begins 52 when photosystems within the symbiont cells become damaged by heat and sunlight and cells are 53 subsequently ejected by the coral host (Jones et al. 1998, Warner et al. 1999). In addition to 54 symbiont-related mechanisms of coral bleaching, bleaching can be a physiological response of the 55 coral, in which case genetic variation among coral could affect their response. Some evidence 56 exists for this mechanism. When experimentally exposed to warm water, populations of Porites 57 astreoides from different temperature conditions (no more than 10km apart) showed different 58 bleaching responses despite harboring the same Symbiodinium communities. These responses 59 were associated with differences in gene expression and significant genetic divergence correlated 60 with in situ temperature conditions (Kenkel et al. 2013, Kenkel andMatz 2016). A third 61 mechanism for coral bleaching is the probiotic hypothesis (Reshef et al. 2006). This mechanism 62 has highlighted the importance of microbial communities in coral mucus and tissues that change 63 in response to abiotic conditions such as temperature (Bourne et al. 2008, Li et al. 2015. Studies 64 have shown that increasing water temperature is associated with a shift in bacterial community 65 compositions and virulence patterns and that following temperature stress, bacterial communities 66 slowly return to their original state (Bourne et al. 2008, Rosenberg et al. 2009). 67 These mechanisms are usually studied separately and do not consider the effect of 68 population dynamics of the coral host. This oversight may be partly due to the difficulty of 69 studying population genetics in many coral genera until the recent application of restriction-site Islands. Montipora are generalists in their Symbiodinium community composition but are generally 79 more sensitive to environmental conditions than Porites, which are largely inflexible to shifting 80 symbiont composition (Putnam et al. 2012). Growth rates differ between the taxa, with Montipora 81 having high growth rates and Porites a comparatively low rate (Gladfelter et al. 1978, Huston 82 1985. This suggests that there may be an inherent fitness tradeoff associated with symbiont 83 switching ability. Montipora switch symbionts to optimize for fast growth at the expense of 84 environmental sensitivity while Porites exhibit high symbiont fidelity that confers environmental 85 resilience but slower growth. Because of this inherent difference, it is imperative for the field to 86 better understand if these corals, with fundamentally different life history strategies, differ in their 87 genetic structure.

88
Kaneohe Bay is a well-studied marine system that is uniquely positioned to explore these 89 questions. The Hawaii Institute of Marine Biology (HIMB) sits upon Coconut Island in the 90 southern, sheltered portion of the bay and is the gateway for much of the research that comes out 91 of the bay. As a result of this, episodes of extreme stress, like heatwaves and freshwater kills, are 92 well-documented and the patterns of bleaching in 1996 and 2014 documented by researchers at 93 HIMB provide some context for this present study Brown 2004, Bahr et al. 2015a, 94 2017). In addition to its recent temperature-related stressors, the bay has a long history of human 95 utilization that began with Polynesian settlement and has more recently been subject to invasive 96 species introduction, agricultural runoff, sewage discharge, and extensive dredge and fill 97 operations (Bahr et al. 2015b). In an otherwise well-studied system, the bay is understudied in 98 regards to the population genetics of their hallmark organisms: corals. This study sought to fill this    duplicated 10x to increase "read" depth, converted from fasta to fastq using dummy quality scores, 161 and included in the assembly. No in-silico data was needed for P. compressa as ipyrad allows for 162 the inclusion of reference genotypes (P. lutea reference genome) as a sample in output files.

163
Following quality checks, data for each species was assembled using the ipyrad 0.   temperature range, monthly salinity range, weekly salinity range, and daily salinity range at an 231 alpha level of p<0.05 (Fig. 2).  Harvester (Earl and vonHoldt 2012) indicated the optimal value of K for Montipora samples to be 240 K=2 while the optimal value of K for Porites samples was found to be K=3 (Supplemental Figure   241   1). Site-level and individual-level probability of membership for each species are shown in Fig. 3. 242 The STRUCTURE analyses reveal clear population structure patterns in Montipora but no   two clonal triplets and four clonal pairs were detected (Fig. 5). Spatially, these clonal groupings 265 occurred predominantly at outer bay sites 2, 4, 6, and 8, with only one inner bay colony, P3W_A, 266 being represented as part of a clonal group. Clonal colonies made up the majority of samples 267 recovered in sites 2, 4, and 6. At these three sites, a total of 17 genotypes were expected but only 268 11 were detected using our sampling design and ddRAD methods. Padilla-Gamiño and Gates 2012) (Fig. 6). Because water cannot easily escape the sheltered 319 compressa and P. lobata . It is plausible that this strongly-supported basal 320 clade is present due to cryptic species or hybridization and introgression between species.

321
Additional evidence of cryptic species or reticulate evolution can be found in distributions of 322 percent pairwise similarity between P. compressa individuals (Fig. 5) immediately adjacent to our sites 1, 2, 3, 5, and 6. In this set of surveys, it was found that sites in spatially. Additionally, it was found that recovery rates increased the further north individuals were 370 within the bay. Our study shows that there is population structure along a north-south gradient 371 within KB and that this aligns with the spatial distribution of post-bleaching recovery rates.

372
It is important to note that these bleaching events were fundamentally different, as