资源环境科技发展态势分析平台

Soil respiration (Rs) is the largest terrestrial carbon (C) efflux to the atmosphere and is predicted to increase drastically through global warming. However, the responses of Rs to global warming are complicated by the fact that terrestrial plant growth and the subsequent input of plant litter to soil are also altered by ongoing climate change and human activities. Despite a number of experiments established in various ecosystems around the world, it remains a challenge to predict the magnitude and direction of changes in Rs and its temperature sensitivity (Q(10)) due to litter alteration. We present a meta-analysis of 100 published studies to examine the responses of Rs and Q(10) to manipulated aboveground and belowground litter alterations. We found that 100% aboveground litter addition (double litter) increased Rs by 26.1% (95% confident intervals, 18.4%-33.7%), whereas 100% aboveground litter removal, root removal and litter + root removal reduced Rs by 22.8% (18.5%-27.1%), 34.1% (27.2%-40.9%) and 43.4% (36.6%-50.2%) respectively. Moreover, the effects of aboveground double litter and litter removal on Rs increased with experimental duration, but not those of root removal. Aboveground litter removal marginally increased Q(10) by 6.2% (0.2%-12.3%) because of the higher temperature sensitivity of stable C substrate than fresh litter. Estimated from the studies that simultaneously tested the responses of Rs to aboveground litter addition and removal and assuming negligible changes in root-derived Rs, "priming effect" on average accounted for 7.3% (0.6%-14.0%) of Rs and increased over time. Across the global variation in terrestrial ecosystems, the effects of aboveground litter removal, root removal, litter + root removal on Rs as well as the positive effect of litter removal on Q(10) increased with water availability. Our meta-analysis indicates that priming effects should be considered in predicting Rs to climate change-induced increases in litterfall. Our analysis also highlights the need to incorporate spatial climate gradient in projecting long-term Rs responses to litter alterations.