The list of holobiont characteristics contributing alone or in combination to a coral stress response (Fig. 1) is doubled by the nature of the coral–dinoflagellate symbiosis itself. Indeed, even this is a gross oversimplification because it ignores the diverse and complex microbiota resident on and in corals (Rohwer et al. 2002) that are also critical players in coral health (Rosenberg et al. 2007). It is not surprising, therefore, that there is no unified view held by coral ecologists as to the dominant or primary factors that determine where a coral holobiont falls on the susceptibility–resilience continuum and whether its position could change or shift with climate change.
The most obvious holobiont characteristics potentially contributing to a stress response are the host and symbiont species partnership combinations. Many studies have examined the dynamics and differential performance of the symbiont clades and subclades in corals before, during and after stress. There is clear evidence that some symbionts are more stress tolerant than others (Rowan 2004; Robison & Warner 2006; Loram et al. 2007). Much interest has been focused on clade D symbionts in the IndoPacific as being more stress tolerant than the dominant clade C types (Baker et al. 2004; Rowan 2004; Abrego et al. 2008; Oliver & Palumbi 2009). These studies set up the question of whether an increase in the prevalence of D-containing hosts after bleaching events is a result of differential survival of these D-containing hosts or of shuffling or switching of symbionts types within the host after bleaching (Baker et al. 2004; Berkelmans & van Oppen 2006; Sampayo et al. 2008). Host species also vary in their ability to withstand the effects of bleaching. Differential production of mitigative antioxidants (Richier et al. 2005) and varying abilities to acquire nutrition heterotrophically by feeding when bleached (Grottoli et al. 2006) are two examples in hosts described to date.
There is ample evidence that corals can be tolerant of and even thrive in temperature extremes (Coles & Brown 2003). In addition, there are plenty of examples of physiological response variation in corals, such as differing levels of fluorescent proteins, mycosporine-like amino acids and a diversity of antioxidant enzymes (Baird et al. 2009). However, the big question remains: what are the mechanisms that result in differences in thermal tolerance and physiological response? Are they due to genotypic adaptation—where selection drives differences in susceptibility over evolutionary times scales or rather are they due to phenotypic acclimatization—where organisms respond to extremes using their existing genomic repertoire within the lifetime of an individual? There is evidence of genotypic adaptation of several host species to different thermal environments across broad latitudinal gradients on the Great Barrier Reef (Smith-Keune & van Oppen 2006; Ulstrup et al. 2006). In addition, phenotypic plasticity of both host- and symbiont-specific characters in response thermal stress has been documented. The coral Goniastrea aspera was found to be resistant to hyperthermic-stress-induced bleaching when colonies had previously experienced stress and bleaching (Brown et al. 2002). Likewise, symbionts photoacclimated to high light levels were found to be tolerant of a hyperthermic stress compared with nonacclimated controls (Robison & Warner 2006). It is these questions of adaptation and acclimatization that lie at the very centre of Barshis and coworkers’ study.
The authors performed a reciprocal transplant study of corals between a fore reef and back reef site in American Samoa within 5 km of one another with very different thermal characteristics. The fore reef site was a more stenothermal environment with very low daily fluctuations in temperature. In contrast, the back reef site was a eurythermal environment with daily temperature fluctuations of up to 6 °C. Colonies of P. lobata were transplanted and then sampled, over a period of several days, for genetic identity of both symbiont and host and for a suite of physiological biomarkers for stress in host tissues: heat shock protein 70, manganese superoxide dismutase and ubiquitin-conjugated and 4-hydroxynonenal-conjugated proteins. The investigators found significant fixed and environmental effects on biomarkers response but with high variability on both temporal and treatment scales. This points to the great difficulty in using high-resolution diagnostic tools, that limit sample size because of their labour- and cost-intensive nature, on highly variable field samples (van Oppen & Gates 2006). However, the response of ubiquitin conjugates was one stunning exception: all samples originating from the fore reef stenothermal site, regardless of transplant destination or time of sampling, had very low levels of ubiquitin conjugates compared to back reef samples. These results indicated the presence of a fixed, or genotypic, rather than a plastic, phenotypic, effect on this physiological response.
The real surprise to me came with the analysis of the genetic makeup of hosts and symbionts between the two sites. Whereas the symbionts were found to be a homogenous population across the sites, the authors found significant genetic differentiation in host populations between the two sites. Such subdivision across such a small distance, <5 km, suggests strong selection for physiological adaptations to these differing temperature regimes.
These results expand our thinking about how corals might respond to rapid climate change. P. lobata from the stable fore reef did not upregulate ubiquitin conjugates when placed in the fluctuating back reef. This could indicate that genotype constrains or limits the capacity of corals to respond to thermal changes. Furthermore, the absence of symbiont population subdivision or cladal differences between sites, unlike findings in some other studies (e.g. Oliver & Palumbi 2009) suggests that dynamic changes in symbiont make up with stress may not always occur, further limiting the capacity to adjust to temperature increases. This study therefore expands the possibilities of dynamic contributions and interplay between holobiont characteristics that determine a stress response. Discerning and predicting patterns among these characteristics might ultimately help managers in decisions about which corals or areas of reef to protect as the planet warms.