Viral infection of microbes is probably the most common ecological interaction on the planet. However, it is unknown what fraction of infections are rapidly lethal or establish long-term virus-host symbioses because viruses can act as lytic predators or temperate parasites. Viruses sculpt the evolution, composition, structure, and vital rates of host communities from molecular to global scales and affect ecosystem processes like biogeochemical cycling, productivity, and food webs. While most viral ecologists have focused on these phenomena as an outcome of density-dependent lethal lytic infection, my research shows that these ecosystem processes are more likely to the outcome of physiology-dependent long-term temperate infections. This changes – even upends – the narrative of virus-host interactions from being deterministic and easily measured as host and viral densities to a much more challenging and long-term one centered around host physiology. Importantly, this mechanistic shift also means that accepted outcomes of infection like how and why energy and matter flow through ecosystems must be revised from cellular to global scales.