Soil respiration fluxes in a temperate mixed forest: seasonality and temperature sensitivities differ among microbial and root-rhizosphere respiration.

Although soil respiration, a major CO(2) flux in terrestrial ecosystems, is known to be highly variable with time, the response of its component fluxes to temperature and phenology is less clear. Therefore, we partitioned soil respiration (SR) into microbial (MR) and root-rhizosphere respiration (RR) using small root exclusion treatments in a mixed mountain forest in Switzerland. In addition, fine root respiration (FRR) was determined with measurements of excised roots. RR and FRR were strongly related to each other (R(2) = 0.92, n = 7), with RR contributing about 46% and FRR about 32% to total SR. RR rates increased more strongly with temperature (Q(10) = 3.2) than MR rates (Q(10) = 2.3). Since the contribution of RR to SR was found to be higher during growing (50%) than during dormant periods (40%), we separated the 2-year data set into phenophases. During the growing period of 2007, the temperature sensitivity of RR (Q(10) = 2.5, R(2) = 0.62) was similar to that of MR (Q(10) = 2.2, R(2) = 0.57). However, during the dormant period of 2006/2007, RR was not related to soil temperature (R(2) = 0.44, n.s.), in contrast to MR (Q(10) = 7.2; R(2) = 0.92). To better understand the influence of plant activity on root respiration, we related RR and FRR rates to photosynthetic active radiation (both R(2) = 0.67, n = 7, P = 0.025), suggesting increased root respiration rates during times with high photosynthesis. During foliage green-up in spring 2008, i.e., from bud break to full leaf expansion, RR increased by a factor of 5, while soil temperature increased only by about 5 degrees C, leading to an extraordinary high Q(10) of 10.6; meanwhile, the contribution of RR to SR increased from 29 to 47%. This clearly shows that root respiration and its apparent temperature sensitivity highly depend on plant phenology and thus on canopy assimilation and carbon allocation belowground.