© 2020 John Wiley & Sons Ltd Improving coral reef conservation requires heightened understanding of the mechanisms by which coral cope with changing environmental conditions to maintain optimal health. We used a long-term (10 month) in situ experiment with two phylogenetically diverse scleractinians (Acropora palmata and Porites porites) to test how coral–symbiotic algal interactions changed under real-world conditions that were a priori expected to be beneficial (fish-mediated nutrients) and to be harmful, but non-lethal, for coral (fish + anthropogenic nutrients). Analyzing nine response variables of nutrient stoichiometry and stable isotopes per coral fragment, we found that nutrients from fish positively affected coral growth, and moderate doses of anthropogenic nutrients had no additional effects. While growing, coral maintained homeostasis in their nutrient pools, showing tolerance to the different nutrient regimes. Nonetheless, structural equation models revealed more nuanced relationships, showing that anthropogenic nutrients reduced the diversity of coral–symbiotic algal interactions and caused nutrient and carbon flow to be dominated by the symbiont. Our findings show that nutrient and carbon pathways are fundamentally “rewired” under anthropogenic nutrient regimes in ways that could increase corals’ susceptibility to further stressors. We hypothesize that our experiment captured coral in a previously unrecognized transition state between mutualism and antagonism. These findings highlight a notable parallel between how anthropogenic nutrients promote symbiont dominance with the holobiont, and how they promote macroalgal dominance at the coral reef scale. Our findings suggest more realistic experimental conditions, including studies across gradients of anthropogenic nutrient enrichment as well as the incorporation of varied nutrient and energy pathways, may facilitate conservation efforts to mitigate coral loss.