1. Introduction
Effects of climate change are projected to become more intense during the 21st century (IPCC, 2014; Caldeira & Wickett, 2003). Tropical coral reefs have been most negatively affected by the rapid climate change, resulting in worldwide coral reef decline (Parmesan, 2006; Heron et al., 2016). Consequently, corals must adapt to their changing environments, or face declines and ultimately extinction.
Thermal adaptation has been documented in various coral populations throughout the Indo-Pacific, but it is uncertain whether these disparate populations adapt to other novel environmental stressors in similar manner (Palumbi et al., 2014). If the same genes are under selection by different stressors, evolution of resistance to one stressor would promote the resistance to other stressors, speeding up the adaptation to long-term climate change. It is also possible that the same genes are used in adaptation to different stressors but differentially selected upon, which would imply a tradeoff where some adaptations exclude others. Such genetic tradeoffs could slow down or even entirely prevent long-term adaptation to multiple stressors. Lastly, if different genes are used for adaptation to different stressors, this would imply independent evolution and that these adaptations can occur at the same time, neither helping nor hindering each other.
Previous studies investigating how coral responds to multiple environmental stressors, simultaneously, have done so by exposing different coral genotypes to various stressors and then measuring the coral’s response to each stressor or combination of stressors. The experimental design of Mueller et al., 2018 found that Acropora cervicornis resistance to white-band disease and high-water temperatures had no tradeoffs, and that these responses evolved independently. The elevated temperature, increasedp CO2, and bacterial challenge treatments used in Wright et al., 2019 found that Acropora millepora ’s responses to these treatments had no tradeoffs, and that coral’s survival under one treatment indicated its potential to withstand the other treatments. The physiological responses of Acropora cervicornis to raised temperatures and acidification in Mueller et al., 2021 revealed no tradeoffs between the responses, and that the phenotypic responses had significant variation among the coral genotypes. These previous studies broadened the scope of research on how climate change affects coral reefs to include the effects of multiple environmental stressors, rather than individual, and their interactions (Pendleton et al., 2016). However, coral’s response pathways and outcomes when faced with these multiple climate changes, as well as others, still needs to be addressed (Pendleton et al., 2016). Here, we aim to address this problem by investigating the evolutionary pathways of coral’s adaptive response to various environmental stressors, and ultimately determine if these pathways are the same, independent, or trade off each other.
We reanalyzed transcriptomic reads from four studies comparing corals across four environmental gradients (Figure 1): differentially heated tidal pools (Rose et al., 2017), proximity to CO2 seep (Kenkel et al., 2017), different symbionts (Barfield et al., 2018), and latitudinal gradient (Dixon et al., 2014). We have quantified not only gene expression differences between contrasts, which was the original goal of all these studies, but also genetic divergence based on the same transcriptomic reads. We hypothesized that coral’s divergence and gene expression responses to daily heat stress, less heat-tolerant type of algal symbionts, and warmer location in the latitudinal gradient should show at least some parallelism, both at the level of genetic divergence and gene expression, as all these are expected to involve thermal adaptation. Even more obvious should be the similarity between the two cases of adaptation and/or acclimatization to a high CO2environment, at Dobu and Upa-Upasina. At the same time, we expected that CO2 adaptation would be either independent, or trading off the thermal adaptation. It is important to note that the tradeoff could be detected only for gene expression data, i.e., opposite regulation of the same genes (Figure 1). Tradeoff detection was not possible for genetic divergence since the test statistic,F ST, does not capture the direction of evolutionary change. Finally, we examined the overlap between differentially expressed and genetically diverging genes, to look for evidence that the regulatory differences had genetic basis.
Figure 1. Experimental design of the study. (a) Environmental stressors, showing stressed and non-stressed contrast, selected for study. (b) Contrast’s non-stressed and stressed populations were used to compute the per gene genetic divergence (F ST) and gene expression. (c) To compare directions of adaptive responses to these stressors, a principal coordinate analysis (PCoA) and principal component analysis (PCA) was used for F ST (with Bray-Curtis distance measure) and log-fold change, respectively. (d) Possible outcomes of the PCoA or PCA include adaptive response vectors aligning themselves together to reflect parallel evolution, aligning themselves orthogonally (i.e., 90°) to reflect independent evolution, or aligning themselves 180° apart to reflect tradeoffs between the responses (possible for gene expression analysis).