Abstract: 【Aim】 The camellia weevil, Curculio chinensis Chevrolat, is an important pest attacking fruits of oil tea Camellia, an endemic genus to China, and causes tremendous fruit drop of its host plants. In the study area, the weevil larvae and its host (Camellia meiocarpa) fruits are strongly preyed by a rodent. To provide theoretical basis for scientific prevention and control of the weevil pest, we explored the complicated interactions among Camellia meiocarpa, Curculio chinensis and the rodent, and their behavioral mechanisms. 【Methods】 During the fruit drop season, 6 161 dropped fruits of C. meiocarpa were collected systematically from a farm in Heshan District, Yiyang City, Hunan Province, the damaged situation of each fruit caused by the weevil C. chinensis or the rodent was recorded, and the length, diameter and pericarp thickness of each fruit were also measured. The relationships between the weevil infestation rate and fruit traits such as the length, diameter and pericarp thickness of fruits as well as the correlation between the weevil mortality and the survey date were analyzed through linear regression, and the difference in fruit size between weevil-infested and non-weevil-infested fruits preyed by rodents was analyzed by independent samples t test. 【Results】 The weevil-infested fruit drop started on July 2 and lasted till fruits were harvested, and the peak of weevil-infested fruit drop was from August 16 to August 23 in 2013. Rodent predation was throughout the fruit drop season, and its peak was as the same as that of the weevil-infested fruit drop. The percentage of the 4th-5th instar larvae to the total number of weevil larvae in dropped fruits was 85.5%. There was no correlation between the weevil infestation rate and fruit diameter, and the weevil infestation rate had a significantly negative correlation with the pericarp thickness, but a significantly positive correlation with the benefit-cost ratio. The earlier the weevil infested fruits dropped, the heavier the fruit development was restricted, and the higher the weevil larval mortality was. The proportion of smaller fruits to the total number of rodent-preyed fruits was 76.9%, and all seeds in smaller fruits were depleted. The predation rate on weevil-infested fruits was 24.0%, and in these fruits, weevil-infested seeds were almost damaged by rodents, but most of non-weevil-infested seeds were not preyed. Of the rodent-preyed fruits, weevil-infested fruits were significantly larger than non-weevil-infested fruits in terms of fruit length and diameter. 【Conclusion】 The temporal pattern of fruit drop, weevil-infested fruit drop and rodent predation of C. meiocarpa fruits were investigated in this study. The results do not support the hypothesis of active defense, and the primary cause of the early fruit drop is a breakdown in the normal physiological process of fruit development caused by the feeding of C. chinensis larvae. There may be a trade-off between fruit size and pericarp thickness, the two contradictory selective pressures during the oviposition of female adults. In fact, this trade off is that between the fertility of parental females and the fitness of offspring larvae. The temporal pattern of oviposition of parental female adults will ultimately affect the fitness of their offspring larvae through affecting fruit development of their host plant C. meiocarpa. In the study area the rodent is an important natural enemy for C. chinensis, and it can accurately detect and prey on the weevil larvae in bigger fruits.

Key words: Camellia meiocarpa, Curculio chinensis, rodent, weevil infestation rate, predation rate, benefit-cost ratio, fecundity, fitness