Measurements of CO2 flux, soil temperature, and moisture content of selected natural and disturbed soils in central California were made on a monthly basis from August 1994 to October 1995 in an attempt to detect the effects of temperature, moisture, and land use change on CO2 production in soils. Soil CO2 flux displayed a strong negative correlation with soil temperature and a positive correlation with soil moisture at the natural site. However, at the disturbed site, the linear correlation between CO2 flux and temperature/moisture was insignificant. The negative correlation between soil CO2 flux and soil temperature is in contrast to what has been observed in other ecosystems but is typical for Mediterranean ecosystems in which grasses are biologically active only during cool months. Comparison of carbon (C) inventories of paired natural and disturbed soils indicates that both cultivation and logging have resulted in a significant decrease in total soil C content. The reduction in soil C storage is about 26% for the cultivated soil and around 30% for the logged soil. Most of the C loss is from the upper horizons. Radiocarbon (14C) measurements of both recent and archived soil samples demonstrate large differences in C input rate and turnover time between natural and disturbed soils. The average turnover times of organic matter are longer in disturbed soils than in the corresponding natural soils as a result of preferential loss of C from active soil C pools. In both natural and disturbed soils, the average turnover times of organic matter increase with depth from decades or less in shallow horizons to hundreds of years or even thousands of years in deeper horizons. Our results show that land use change can have significant impact on soil C cycle and that shallow soil horizons are most susceptible to disturbance because of shorter turnover times of organic C in these horizons. ¿ 1999 American Geophysical Union