Conservation Tillage Trade‐Offs Are Real but Context‐Dependent: A Global Perspective on Subsoil Management

We thank Kan et al. (2026) for their Letter on our paper (Zhang et al. 2026). Their 15-year conservation deep tillage (CDT) experiment （see also Kan et al. 2025）provides insights into alleviating subsoil constraints. Here we clarify that CDT is promising but largely context-dependent, as illustrated by data comparisons between our global synthesis and their site-specific results. Kan et al. reported results from one long-term field experiment at a single site (Nanjing, China) with a specific soil type. Their yield data consist of only 27 observations for wheat and 33 observations for rice, all from that one location. By contrast, our global synthesis was a second order meta-analysis of 69 published meta-analyses, representing thousands of paired comparisons across diverse climates, soil types, and management systems worldwide. A successful case at one site does not invalidate the statistically robust trade-offs identified globally. Instead, it underscores the need for context-specific innovation, a point we explicitly emphasized in our original paper. This need is further supported by our recent work, which demonstrated that the effectiveness of diversification strategies depends on local conditions (Qiu et al. 2024). Our analysis revealed differential effects of conservation tillage (CT) on ecosystem services versus yield. Table 1 contrasts global CT effects with CDT outcomes. First, the reported CDT effect on soil organic carbon (SOC) is +53%, over 2.5 times the global CT mean (+21%). This suggests either site-specific exceptional conditions or the combined effect of deep tillage and high straw inputs. Our global meta-analyses showed that crop residue return increases SOC (Liu, Fang, et al. 2023) and sustains stoichiometric balance (Liu, Qiu, et al. 2023). Second, while our synthesis detected no significant yield change under CT, Kan et al. reported yield gains (wheat +16%, rice +13%). Third, Kan et al. did not report greenhouse gas (GHG) fluxes or erosion data. In our original study, 41% of paired effects showed increased ecosystem services but decreased yield, with the largest trade-off between yield and GHG mitigation. Thus, the absence of GHG data leaves unresolved whether CDT exacerbates these trade-offs. Kan et al. argue that subsoil compaction limits yield under long-term no-till. We acknowledged subsoil constraints in our original paper and called for innovations like ditch-buried straw return. The CDT data provide a valuable example supporting this direction. Table 1 compares site-specific versus global conditions, suggesting CDT is most effective in high-input, humid systems, not the semi-arid, low-input contexts where CT faces its greatest challenges. Multi-site, multi-year trials are needed before CDT can be recommended globally. The findings of Kan et al. complement rather than refute our synthesis. Our main conclusions remain supported. At the global scale, CT enhances multiple ecosystem services with no net yield change, though partial trade-offs exist, especially between yield and GHG mitigation. Kan et al. take an important step toward context-specific innovations. Further multi-location trials are needed to assess CDT's generalizability. Linchuan Fang drafted the response, contributed to the revision, and approved the final version. This study was supported by the National Natural Science Foundation of China (42477326). The author declares no conflicts of interest. This article is a Response to a Letter to the Editor by (Kan et al, https://doi.org/10.1111/gcb.70942) regarding (Zhang et al, https://doi.org/10.1111/gcb.70825). No new data were generated for this response. All data from Zhang et al. (2026) are available at https://doi.org/10.6084/m9.figshare.30866357.