Energy conservation is a clear function of torpor. Although many studies imply that torpor is also a water-saving strategy, the experimental evidence linking water availability with torpor is inconclusive. We tested the relative roles of water and energy shortages in driving torpor, using the Siberian hamster Phodopus sungorus as a model species. To account for the seasonal development of spontaneous heterothermy, we used male hamsters acclimated to short (8L:16D, SP; n = 40) and long (16L:8D, LP; n = 36) photoperiods. We continuously measured body temperature (Tb) during consecutive 32 h of complete removal of water, food, or both, separated by 7.5 d recovery periods. We predicted that all deprivation types would increase the frequency of spontaneous torpor in SP, and induce torpor in LP-acclimated hamsters. Individuals underwent each deprivation type twice in random orders. Food and water deprivation did not induce torpor in LP-acclimated P. sungorus. Patterns of torpor expression varied among deprivation types in SP individuals. Torpor frequency was significantly lower, but bouts were ∼2 h longer and 2.5 °C deeper, during water deprivation compared to food and food-and-water deprivation. Heterothermic responses to all deprivation types were repeatable among individuals. Different torpor patterns during water and food deprivation suggest that water and energy shortages are distinct physiological challenges. Deeper and longer bouts during water deprivation likely led to higher energy and water savings, while shorter and shallower bouts during fasting may reflect a trade-off between energy conservation and food-seeking activity. The lack of a difference between food- and food-and-water-deprived hamsters suggests a higher sensitivity to food than water shortage. This supports the traditional view that energy conservation is the major function of torpor, but suggests that water shortages may also modulate torpor use. The high repeatability of thermoregulatory responses to resource deprivation suggests that these may be heritable traits subject to natural selection.

1. Three experiments were conducted to investigate the diurnal variation of blood viscosity in broilers. In experiment 1 food and water were supplied freely at 20 degrees C (20-FW). In experiment 2 food and water were withdrawn at 20 degrees C (20-NFW), while in experiment 3 food and water withdrawn at 30 degrees C (30-NFW). 2. Blood sampling time points were 09.00 h, 15.00 h, 21.00 h, 03.00 h and 09.00 h the next day in each experiment. 3. In all experiments, whole blood viscosity (WBV), red blood cell count (RBC) and haematocrit (HCT) were greater during the dark (21.00 h and 03.00 h) than during the light period. During the dark period, there were no differences in WBV, RBC and HCT between 20-FW and 20-NFW, or between 20-NFW and 30-NFW. At 09.00 h, WBV and HCT were higher in 20-FW than in 20-NFW. At 15.00 h and 09.00 h (day 2), WBV and HCT were greater in 20-NFW than in 30-NFW. 4. There were no light-dark differences in plasma viscosity (PV), plasma protein concentration (PPC) or mean corpuscular volume (MCV) in any experiment. However, 20-NFW birds had a lower PPC and higher MCV compared with 20-FW, and a higher PPC and lower MCV compared with 30-NFW, while no difference was found in PV. 5. WBV increased linearly with RBC and HCT. PV increased with PPC, while MCV decreased. 6. These results indicate that there is diurnal variation in whole blood viscosity, which is greater during the dark than during the light period. During the light period it is strongly influenced by high environmental temperature and food and water withdrawal.