不同施肥下中国旱地土壤有机碳变化特征——基于定位试验数据的Meta分析

搜集1994—2011年国内外有关中国旱地施肥处理的102个定位试验点的1146对田间试验数据,采用Meta-analysis方法定量分析了不同施肥条件下我国旱地耕层土壤有机碳(Soil Organic Carbon,SOC)的变化特征。结果表明,与对照(CK)相比,不同施肥措施均能显著提高耕层SOC含量,但不同施肥措施的效应不同。氮磷钾肥配施有机肥处理下SOC增速最大,为0.38 g kg-1a-1,单施磷肥处理增速最小,SOC增速仅为0.032 g kg-1a-1;添加有机肥的处理SOC增速远大于仅有无机化肥投入的施肥处理。不同施肥处理下SOC增速存在一定的空间分异特征且不同时期试验SOC相对变化速率也不相同,早期试验中SOC增速大于中、后期试验;不同种植制度对SOC变化速率的影响亦不同,有机肥的投入可以降低种植制度对SOC变化的影响。SOC积累与否及其幅度并不完全取决于初始SOC含量;随试验年限的增加,SOC增加速率呈降低趋势,仅采用短期试验(11 a)数据可能高估施肥措施下的固碳潜力。

A meta-analysis of field experiment data for characterizing the topsoil organic carbon changes under different fertilization treatments in uplands of China

Soil organic carbon (SOC) pool is the largest carbon (C) reservoir in the terrestrial biosphere, and SOC storage in agricultural soils is extremely sensitive to the management practices such as tillage, fertilization, and crop residue management. Consequently, quantifying the changes in SOC resulting from the implementation of agricultural management practices is important for improving soil fertility, maintaining crop yield, and offsetting the emission of carbon dioxide. A total of 1146 paired-treatment data derived from the 102 published fertilization experiments (1994-2011) and meta-analysis method were employed in this study for characterizing the changes in topsoil organic carbon contents under different fertilization treatments in uplands of China. In order to identify the spatiotemporal changes in SOC and the variations of SOC change resulting from cropping system, initial SOC content, and duration of the experiments, meta-analysis was conducted by using the Meta package of R software (a language and environment for statistical computing and graphics), and then the changes in SOC contents were analyzed by taking the relative change rate (RC) of SOC as an appropriate effect size estimator in the analysis so that the treatment and control means of the experiments can be compared directly. Results showed that all fertilization treatments significantly sequestrated C in soils when compared with the control treatment. The RC of SOC under NPKM (mineral nitrogen, phosphorus, potassium, and manure) treatment was the largest, with a rate of 0.38 g kg^-1a^-1, while the P (mineral phosphorus only) treatment caused the lowest RC with 0.032g kg^-1a^-1. The RCs of SOC under fertilization treatments with manure incorporation were obviously higher than those treated only by inorganic chemical fertilizers. The regional characters of RCs under manure or inorganic fertilizer treatments, presenting as a magnitude order, can be identified as South > North China (NC) > Northwest (NW) > Northeast (NE), but such magnitude order changed when manure and inorganic fertilizers were applied simultaneously. Moreover, the RCs of SOC derived from the experimental data that observed in the early stage of the experiments (before 1990) were obviously higher than those obtained in the middle (1991-2000) or later (2001-2010) stage of the experiment. Furthermore, the accumulating effects of SOC resulting from fertilization treatments were also different for single and double cropping system. In the situation of joint use of manure and inorganic fertilizer, the RCs of SOC under single-season cropping system were higher than those under double-season cropping system, otherwise, an opposite trend in SOC change can be observed. The application of manure may diminish the variations of changes in SOC for different cropping systems, but the impacts of initial SOC content on the RCs of SOC were unclear. The SOC sequestration potential may be overestimated when only short-term experiment (duration <11a) data were used for the estimation, because the experimental duration time had significant influences on the RCs of SOC under different fertilization treatments.