Soil water content and temperature as independent or confounded factors controlling soil respiration in a temperate mixed hardwood forest

Variation in soil temperature can account for most of the seasonal and diel variation in soil CO2 efflux, but the temperature effect is not always consistent, and other factors such as soil water content are known to influence soil respiration. The objectives of this research were to study the spatial and temporal variation in soil respiration in a temperate forested landscape and to evaluate temperature and soil water functions as predictors of soil respiration. Soil CO2 auxes were measured with chambers throughout an annual cycle in six study areas at the Harvard Forest in central Massachusetts that include soil drainage classes from well drained to very poorly drained. The mean annual estimate of soil CO2 efflux was 7.2 Mg ha(-1), but ranged from 5.3 in the swamp site to 8.5 in a well-drained site, indicating that landscape heterogeneity is related to soil drainage class. An exponential function relating CO2 fluxes to soil temperature accounted for 80% of the seasonal variation in fluxes across all sites (Q(10) = 3.9), but the Q(10) ranged from 3.4 to 5.6 for the individual study sites. A significant drought in 1995 caused rapid declines in soil respiration rates in August and September in five of the six sites (a swamp site was the exception). This decline in CO2 fluxes correlated exponentially with decreasing soil matric potential, indicating a mechanistic effect of drought stress. At moderate to high water contents, however, soil water content was negatively correlated with soil temperature, which precluded distinguishing between the effects of these two confounded factors on CO2 flux. Occurrence of high Q(10) values and variation in Q(10) values among sites may be related to: (i) confounding effects of high soil water content; (ii) seasonal and diel patterns in root respiration and turnover of fine roots that are linked to above ground phenology and metabolism; and (iii) variation in the depth where CO2 is produced. The Q(10) function can yield reasonably good predictions of annual fluxes of CO2, but it is a simplification that masks responses of root and microbial processes to variation in temperature and water content throughout the soil.