On the “strict homeostasis” assumption in ecological stoichiometry

► We present two closed-nutrient producer-consumer interaction models. ► We discuss when the widely used “stoichiometric homeostasis” assumption is valid. ► Strict homeostasis is reasonable when the stoichiometric variability of the consumer is less than a threshold. ► Strict homeostasis seems valid for many herbivores except for herbivores with small mortality rates. ► The two-nutrient model shows how the threshold shifts according to the change of environmental N:P ratio.