The allometry of energy reserve depletion: test of a mechanism for size-dependent winter mortality.

We experimentally tested the hypothesis that energy reserve depletion varies inversely with size in the fish Menidia menidia, an estuarine fish known to exhibit size-dependent winter mortality. Individuals in two size groups were starved at two winter temperatures (4°and 8°C) and sacrificed at a range of time intervals (up to 127 days). Lipid levels and lean tissue were analyzed to estimate somatic energy storage. As predicted, energy depletion was greater at high temperatures, and proportionally greater in small than in large fish. After 60 days of starvation at 4°C, small fish retained an average of 67% of their original energy reserves (vs 53% at 8°C), while large fish retained an average of 80% (vs 66% at 8°C). At 4°C, fish that were fed depleted their energy reserves as rapidly as unfed fish, but at 8°C, fish that were fed maintained reserves at higher levels than unfed fish. A high proportion of unfed fish (56% at 4°C, 27% at 8°C) died before they were to be sacrificed. Survival probability did not vary with size, nor was it influenced by the amount of energy reserves. The rate of energy depletion (equivalent to routine metabolic rate) decreased gradually over time, particularly in small fish. Routine metabolism did not conform to a single scaling relationship. Within each temperature-size group, the routine rate declined more rapidly than metabolically active mass (lean mass). At 8°C, the difference between size groups in energy depletion rate conformed closely to the expected allometry exponent of 0.8. In contrast, at 4°C, the estimated allometry exponent increased over the experiment (-0.19 to 2.5). We conclude that strategies to minimize energy loss may often modify bioenergetic scaling relationships.