FCR | 欧洲保护性农业中的谷物产量、产量稳定性和一氧化二氮释放:Meta分析

Fig. 1. The geographical positions of the 58 field experiments on cereal yields and 11 field experiments on N2O emissions included in the meta-analysis, mapped onto the environmental zones of Europe as defined by Metzger et al. (2005).

Fig. 2. The distribution of (a) the relative yield, ln (R), and (b) the relative coefficient of temporal yield variation, ln (CVR), functioning as an inverse measure of yield stability. The diamond indicates the weighted overall effect with bias-corrected 95 % CIs. The dashed vertical line indicates conventional agriculture (CONV), and n is the number of independent studies.

Fig. 3. Forest plots presenting for each individual study (a) the relative yield, ln (R), and (b) the relative coefficient of temporal yield variation, ln (CVR), functioning as an inverse measure of yield stability. The black squares indicate mean effect sizes, with 95 % CIs. The square size corresponds to study weight. The diamond indicates the weighted overall effect with bias-corrected 95 % CIs. The dashed vertical line indicates conventional agriculture (CONV).

Fig. 4. Combined effects of (a) tillage depth in conservation agriculture and topsoil clay content and (b) tillage depth and mineral N fertilizer on the relative yield (ln R). The doses of mineral N fertilizer are similar in conventional agriculture (CONV) and conservation agriculture. Symbol size corresponds to study weight. The dashed horizontal line indicates CONV. Qm, model heterogeneity, and n is the number of independent studies. Note that the clay content was documented in 54 studies, whereas the quantity of N fertilizer was reported in 53 studies.

Fig. 5. Percentage change in (a) cereal yields and (b) the coefficient of temporal yield variation, functioning as an inverse measure of yield stability, due to the implementation of conservation agriculture (CA) in comparison to conventional agriculture (CONV) across different cereal types (winter wheat, spring and winter barley and maize). The dashed vertical line indicates CONV. The numbers in parentheses indicate the number of independent studies. The changes in yield or its temporal variation were considered significantly different from the CONV if the bias-corrected 95 % CIs did not overlap with zero. Qb, between-group heterogeneity model, and n is the number of independent studies.

Fig. 6. Predicted change in cereal yield across Europe under two conservation agriculture scenarios: (a) no-tillage and (b) minimum tillage (15 cm depth), both adhering to the core principles of CA. Predictions are from a spatialized General Linear Model (GLM), and changes are expressed as a percentage relative to conventional agriculture.

Fig. 7. (a) Forest plot showing relative N2O emissions (ln RN2O) for each study and subgroup overall for soil texture groups. The squares indicate mean effect sizes, with 95 % CIs. The diamond square indicates the weighted overall effect with bias-corrected 95 % CIs. Asterisks indicate statistically significant effects of conservation agriculture at p < 0.05. (b) Weighted meta-regression between clay content and effect size (ln RN2O). Qm, model heterogeneity, and n is the number of independent studies. Cross symbol indicates outlier (Baggs et al., 2003). The dashed lines indicate conventional agriculture (CONV). The symbol size corresponds to study weight.

Fig. 8. A diagram illustrating the main results of this meta-analysis on the effect of conservation agriculture on (a) cereal yield, (b) yield stability, and (c) soil N2O emissions when compared to conventional agricultural practices. MT, minimum tillage, NT no-tillage, both adhering to the core principles of conservation agriculture (CA). Key moderating factors are highlighted in brown boxes.