Abstract: The predation risk allocation hypothesis states that predation risk has an effect on animal’s behavior， and the animal will make trade-offs between vigilance and other behaviors such as foraging and storing according to the level of predation risk. When the risk is high， animal will reduce foraging and other behaviors， and allocate more time on vigilance， and vice versa. This hypothesis has been proved to be true for some animals such as Ochotona curzoniae， but for the wide distributed species in deserts of Central Asian Rhombomys opimus， this hypothesis has not been tested yet. This paper used the focal animal sampling method to get the behavior data of the [R. opimus，]and tested the predation risk allocation hypothesis. The research was conducted during the breeding season of the R. opimus from May to August in the year of 2010. The R. opimus’s behavior was observed from 8:00 to 22:00 everyday. The behaviors of the R. opimus were classified into 4 kinds: vigilance， foraging， storing， and other activities. And the paper treated all 4 kinds of these activities as the total ground activities. Each behavior was recorded by total duration time， mean duration time of each time， and frequency of the activities for each 60 min observation sessions. The paper used Pearson Correlations to detect the relationships between vigilance and foraging， storing， total ground activities and other behaviors. The results indicated that vigilance has a significant negative relationship with foraging and other behaviors compared by total duration time in each 60 min observation samples （vigilance vs total ground activities：r=-0.727，P<0.001；vigilance vs foraging：r=-0.515，P<0.001；vigilance vs storing：r=-0.174，P<0.001；vigilance vs other behaviors：r=-0.142，P=0.001） and mean duration of each times （vigilance vs total ground activities：r=-0.157，P<0.001；vigilance vs foraging：r=-0.126，P=0.002；vigilance vs storing：r=-0.094，P=0.024）. When compared by activities frequency， vigilance showed a significant positive relationship with foraging and other behaviors （vigilance vs ground activities：r=0.404，P<0.001；vigilance vs foraging：r=0.311，P<0.001；vigilance vs storing：r=0.297，P<0.001；vigilance vs other behaviors：r=0.366，P<0.001）. The results supported the predation risk allocation hypothesis for it showed the pattern of the trade-offs between vigilance and other behaviors. And it indicated that predation risk allocation hypothesis could be a general hypothesis which states the effects of predator risk on animal’s behavior. Also，the results showed that when [R. opimus] are more active， they are more vigilant. This results indicated that R. opimus is very vigilant， and vigilance is one of the most frequently and important activities for this species.

Key words: Rhombomys opimus, predation risk, activity time budgets, ground activities, trade-offs