Spatial variation in CO2 efflux from the soil in a mature beech forest ecosystem.

 Soil respiration (Rs) is the second largest carbon flux between the atmosphere and terrestrial ecosystem. Because of the large proportion, even small change in Rs would considerably impact the global carbon cycle. Therefore, it is important to accurately estimate Rs by taking its spatial and temporal variation into consideration. While the temporal variation of Rs and its controlling factors have been well-described, large unexplainable part still has been remained in the spatial variation of Rs especially in the forest ecosystems with complex structures. The objective of this study is to fill the knowledge gap about spatial variation of Rs and its controlling factors in a typical mature beech forest in Japan. Hypotheses of this study were, 1) Rs would show large spatial variation in the mature beech forest, 2) the spatial variation of Rs was mainly influenced by soil water content (SWC) and soil temperature (ST), 3) the two key factors were determined by the forest structures. This study was conducted in a 1- ha permanent study plot in the mature beech forest with significant gap-mosaic structures. To examine these hypotheses, Rs, SWC, ST and parameters related to forest structure, i.e. sum of basal area, diameter at breast height, number of trees, number of species within a radius of 5 m from the Rs measurement points, and canopy openness were measured at 121 points in different season between 2012 to 2013. In this study, all the measurements of Rs were conducted by using alkali-absorption technique.

 Coefficient of variation of Rs was between 25 - 28 % which was similar to that of SWC in all the measurements. The spatial variation of Rs was relatively higher in July, August and September than that in June and October. There was no significant relationship in the spatial variation between Rs and ST in all the measurements, meanwhile, Rs was well explained by SWC in measurements conducted in August, September and October. Multiple linear regression analysis indicated that canopy openness and sum of basal area showed significant positive and negative correlation with SWC, respectively. And canopy openness explained SWC much more than sum of basal area did. This result suggested that SWC, the key factor determined the spatial variation of Rs, cannot be only explained by stems distribution and their characteristics, but also canopy architecture in the forest ecosystem.