Introduction
Coral reefs globally are undergoing significant degradation due to mass coral mortality driven by high sea surface temperatures (Heron et al., 2017; Hughes et al., 2018). Increasingly warm seawater temperatures and losses in coral cover are also leading to changes in fundamental reproductive processes (Baums et al., 2019; Hughes et al., 2019; Shlesinger & Loya, 2019), with major implications for ecosystem resilience and recovery in the aftermath of severe, acute events such as mass coral bleaching. To safeguard the persistence and resilience of coral reefs, assisted evolution interventions to increase the heat tolerance of corals at the early life-history stages as well as in adults are being tested in the laboratory and the field (National Academies of Sciences and Medicine, 2019; Reef Restoration and Adaptation Program: Intervention Technical 2019). These interventions may also protect populations of high importance. Assisted evolution interventions include selective breeding within species, hybridisation between species, and microbial manipulations (Chakravarti, Beltran, & Oppen, 2017; Chan, Peplow, Menéndez, Hoffmann, & van Oppen, 2019; Damjanovic, Blackall, Webster, & Oppen, 2017; Morgans, Hung, Bourne, & Quigley, 2019; Quigley, Bay, & van Oppen, 2019; van Oppen, Oliver, Putnam, & Gates, 2015). The knowledge gained from experimentation of genetic interventions will facilitate optimisation of breeding designs, reduce the incidences and impacts of trait trade-offs and the help to determine the probabilities of success and timeframes associated with genetic gains.
Improved forecasting of future coral reef health based on the scope and rates of recovery is urgently needed. This will rely on estimating rates of acclimatisation and adaptation to predict how populations will respond to a suite of selective pressures. The incorporation of the effects of selection and adaptation to understand population evolutionary trajectories has been applied to the conservation of species like Eltham’s butterfly and bighorn sheep (Creech et al., 2017; Roitman et al., 2017). Corals exist close to their thermal limits (reviewed in Drury, 2019) but exhibit a wide range of phenotypes in response to heat stress (Ainsworth et al., 2015; Kenkel et al., 2013), suggestive of a high adaptive capacity predicated on genetic diversity across populations and species (Matz, Treml, Aglyamova, & Bay, 2018). There is evidence that corals have increased their heat tolerance by ~0.5°C in the last decade, which suggests substantial capacity to adapt to increasing temperatures (Sully, Burkepile, Donovan, Hodgson, & van Woesik, 2019). Part of this heat tolerance can be attributed to the diversity and relative abundance of dinoflagellate symbionts (family Symbiodiniaceae) that inhabit coral tissues (Baird, Bhagooli, Ralph, & Takahashi, 2009; reviewed in Quigley, Baker, Coffroth, Willis, & van Oppen, 2018). In particular, changes in symbiont community composition can attribute 1.0-1.5°C of increased tolerance to adults (Berkelmans & van Oppen, 2006) and 26x increased tolerance in juvenile corals relative to warm-warm and cool-warm crosses (Quigley, Randall, van Oppen, & Bay, 2020) and has been found to explain up to 24% of bleaching variability (Mizerek, Baird, & Madin, 2018). The genomic architecture of heat tolerance is often polygenic (Bay & Palumbi, 2015; Jin et al., 2016). Multiple candidate genes of varying effect sizes have been examined, including heat shock proteins and genes involved in immunity (Louis, Bhagooli, Kenkel, Baker, & Dyall, 2017), although some target genes may more broadly be associated with stress tolerance generally and not heat tolerance specifically. Therefore, to understand corals adaptive capacity, comprehensive knowledge is needed concerning the variation in alleles at particular loci and the topology of polymorphic genomic regions. This knowledge can then be used to understand their influence in driving the emergence of different phenotypes that ultimately enable adaptation in populations over ecological and evolutionary time frames.
Evolutionary genetics describes gene frequency changes across populations and over time. The analysis of genetic diversity in wild and artificially-bred populations is central to understanding the rates and potential of adaptation. Shallow and deep whole genome sequencing is now widely available and relatively cost effective, including sequencing whole genome single-nucleotide polymorphisms (SNPs) (Davey et al., 2011; Helyar et al., 2011) and are increasingly illuminating the ecological and evolutionary mechanisms coral use to respond to their environment (Dixon et al., 2015; Fuller et al., 2019; Palumbi, Barshis, Traylor-Knowles, & Bay, 2014). SNPs are particularly valuable for identifying common variants underlying trait variation and can be applied to genetic conservation practices through breeding programs that utilize quantitative genetic principles, breeding values, and narrow and broad-sense heritability (h2, H2) (Visscher, Hill, & Wray, 2008). SNPs are intrinsically linked to the h2 given the proportionality to the product of SNP quantity and effect size (Holland et al., 2019). This marker-assisted selection approach identifies markers ultimately underlying the heritability estimates, although single alleles may explain only a small proportion of the measured heritable variation (Manolio et al., 2009). Specifically, SNPs provide the marker data used as inputs for relatedness models, that when combined with phenotypic data, are used to calculate additive genetic effects and variances and finally, heritability and the potential for selection (O’hara, Cano, Ovaskainen, Teplitsky, & Alho, 2008).
Here we apply high-density genome-wide marker sequencing to samples collected from selectively bred reproductive crosses to elucidate how artificially produced corals may evolve under heat stress. Selective breeding of corals from historically warmer reefs with others sourced from cooler reefs and then exposing juveniles to temperature tolerant symbionts like Durusdinium trenchii significantly improved coral fitness when exposed to 31°C (e.g. increased survival, growth and bleaching tolerance (Quigley, Randall, van Oppen, & Bay, 2020)), even upon inspection of a limited number of crosses produced from only five parents. Here we link those traits measured in the five reproductive crosses to their underlying allele frequency changes and genomic contribution (h2) and identify the putative causative variants separating the five families to explain and potentially link these heritability estimates at both ambient and elevated temperatures to explain the genetic variation in these three quantitative traits.