The 'hologenome theory of evolution', proposed by Rosenberg and colleagues1, suggests that reef corals, and by extrapolation other multicellular organisms, adapt to environmental stress by alterations in their resident microbial communities. Although this is an interesting idea, we strongly disagree with aspects of this theory, specifically those that arise from the omission and flawed interpretation of the coral literature (for reviews of omitted literature see Refs 2, 3).see fig. 1

Figure 1: A mass bleaching event on the Great Barrier Reef in 2006.
figure 1

These events simultaneously affect multiple species over large geographial ranges and are highly correlated with elevated seawater temperatures.

Full size image

The theory is based on the author's observation that Oculina patagonica bleaches in response to Vibrio shiloi infection and it was subsequently suggested that all coral bleaching is the result of bacterial infection1,4. Our main criticism of this hypothesis is that bleaching is a broad stress response in corals, which has been observed even in the absence of the proposed bacterial causal agents1. In fact, there is no experimental evidence to support a role for bacteria in the mass coral-bleaching events that occur over large geographical scales. The bulk of the literature indicates a direct link between mass bleaching and the anomalous temperatures that disrupt photosynthesis in the endosymbiotic dinoflagellates that are harboured by corals5,6,7,8; this is analogous to the bleaching that occurs in higher plants9. However, this overwhelming evidence was completely ignored by Rosenberg and colleagues. In addition, studies from both terrestrial and marine environments have demonstrated that there is increased disease prevalence and severity in stressed populations10, which suggests that a less-fit holobiont has emerged. These observations completely contradict the expected outcomes of the hologenome theory.

Rosenberg and colleagues also over-simplify the highly complex relationship between host and symbiont, and make broad assumptions about coral and invertebrate immunity. For example, they fail to acknowledge the molecular diversity of the endosymbiotic dinoflagellates and the array of responses and adaptive mechanisms that this union provides the holobiont. They also ignore a range of studies that clearly demonstrate the presence of highly conserved acclimatization mechanisms, which are mobilized to accommodate changes in the environment in both coral and endosymbiotic dinoflagellates. These include the induction of heat-shock proteins and antioxidant enzymes, restructuring of the light-harvesting complexes and differential expression of host pigmentation (reviewed in Ref. 11). In addition, the authors imply that invertebrates lack immune memory, which is a feature of acquired immunity in vertebrates. A range of recent studies (reviewed in Refs 12, 13) have demonstrated the presence of immune memory in invertebrates. Further, a recent survey of available cnidarian genome information shows that Toll and Toll-like-receptor homologues14 are present, which strongly suggests that a pathway for immune memory exists in corals.

In conclusion, Rosenberg and colleagues present a hypothesis that is based on a small subset of the available coral and bleaching literature, and have dismissed without comment well-established facts about endosymbiosis and the host. Their microbial-centric view is at odds with the now widely embraced coral holobiont model15. We feel that the lack of consideration of the wide and diverse scientific literature on coral bleaching, coupled with the absence of scientific evidence that supports a role for bacterial populations in driving global patterns of bleaching, combine to raise serious questions as to the relevance of the hologenome theory of evolution.