Accumulating evidence suggests that resource subsidies are temporally variable; e.g., timing, duration and magnitude. However, few studies have empirically examined how temporal attributes of subsidies are important in structuring consumer populations and their communities. In this talk, I introduce empirical examples of the causes and consequences of temporal variation in subsidies in stream and riparian ecosystems. First, large-scale monitoring revealed that phenological diversity of arthropods and their nematomorph parasites synergistically determined the duration (pulsed vs. prolonged within a season) and seasonal timing (early vs. late in a growing season) of the riparian subsidies into streams. A field experiment then revealed that the pulsed subsidy allowed salmonid fish grow more evenly among individuals, which had different cascading effects on benthic invertebrates and leaf breakdown rate compared with the prolonged subsidy. In another field experiment, fish exhibited a higher numerical response to the early subsidy, but not to the late subsidy, with the timing-dependent life history adoptions and its diversity. We further found that phenological diversity of infected emerging aquatic insects, coupled with the two-year life span of a terrestrial definitive host, prolonged the transmission opportunity for the nematomorph parasite between the aquatic and terrestrial hosts. These empirical studies provide useful bases for developing the theory to explore the effects of the temporal attributes of subsidy on coupled ecosystems.