Coral microbiomes are critical to holobiont health and functioning, but the stability of host–microbial interactions is fragile, easily shifting from eubiosis to dysbiosis. The heat-induced breakdown of the symbiosis between the host and its dinoflagellate algae (that is, “bleaching”), is one of the most devastating outcomes for reef ecosystems. Yet, bleaching tolerance has been observed in some coral species. This review provides an overview of the holobiont’s diversity, explores coral thermal tolerance in relation to their associated microorganisms, discusses the hypothesis of adaptive dysbiosis as a mechanism of environmental adaptation, mentions potential solutions to mitigate bleaching, and suggests new research avenues. More specifically, we define coral bleaching as the succession of three holobiont stages, where the microbiota can (i) maintain essential functions for holobiont homeostasis during stress and/or (ii) act as a buffer to mitigate bleaching by favoring the recruitment of thermally tolerant Symbiodiniaceae species (adaptive dysbiosis), and where (iii) environmental stressors exceed the buffering capacity of both microbial and dinoflagellate partners leading to coral death.

Energy flow and nutrient cycling dictate the functional role of organisms in ecosystems. Fishes are key vectors of carbon (C), nitrogen (N) and phosphorus (P) in aquatic systems, and the quantification of elemental fluxes is often achieved by coupling bioenergetics and stoichiometry. While nutrient limitation has been accounted for in several stoichiometric models, there is no current implementation that permits its incorporation into a bioenergetics approach to predict ingestion rates. This may lead to biased estimates of elemental fluxes. Here, we introduce a theoretical framework that combines stoichiometry and bioenergetics with explicit consideration of elemental limitations. We examine varying elemental limitations across different trophic groups and life stages through a case study of three trophically distinct reef fishes. Further, we empirically validate our model using an independent database of measured excretion rates. Our model adequately predicts elemental fluxes in the examined species and reveals species‐ and size‐specific limitations of C, N and P. In line with theoretical predictions, we demonstrate that the herbivore Zebrasoma scopas is limited by N and P, and all three fish species are limited by P in early life stages. Further, we show that failing to account for nutrient limitation can result in a greater than twofold underestimation of ingestion rates, which leads to severely biased excretion rates. Our model improved predictions of ingestion, excretion and egestion rates across all life stages, especially for fishes with diets low in N and/or P. Due to its broad applicability, its reliance on many parameters that are well‐defined and widely accessible, and its straightforward implementation via the accompanying r‐package fishflux, our model provides a user‐friendly path towards a better understanding of ecosystem‐wide nutrient cycling in the aquatic biome. A free Plain Language Summary can be found within the Supporting Information of this article.