While no-till (NT), one of the pillars of conservation agriculture, is widely practiced in upland crops, considerably less research has been conducted in rice systems. While adoption has been limited, numerous field experiments have evaluated NT rice across diverse regions and management contexts yet reported yield responses to key agronomic practices, soil properties, and climatic conditions remain inconsistent and poorly synthesized at a global scale. Our objective was to conduct a global meta-analysis comparing the effects of NT to conventional tillage (CT) on rice yields and to identify the agronomic practices, soil characteristics, and climate factors most strongly associated with yield changes under NT. We performed a global meta-analysis of 115 studies (579 paired observations) comparing rice yields under NT and CT. Yield response ratios (lnRR) were analyzed using mixed-effects meta-analysis to estimate overall effects. Meta-regression and Random Forest modeling were used to assess relative importance of soil properties, management practices, and climatic factors influencing NT performance. Overall, there was a 4.8% yield reduction under NT. The key factors influencing yield were soil organic carbon (SOC), nitrogen (N) application rate, soil pH, and soil clay content. Soil SOC levels above 18 g ha⁻¹ , N rates above 120 kg ha⁻¹ , a pH range of 6 – 7, and clay content between 180 – 350 g kg⁻¹ were associated with improved NT yields. Residue retention and crop rotation – the other two key components of conservation agriculture along with minimum tillage – had limited influence on yield outcomes in flooded rice systems. This global meta-analysis provides the most up-to-date, and largest, global, quantitative framework for targeting NT. Results indicate that based on soil properties (higher SOC and clay content, moderate pH) coupled with sufficient N inputs, NT yields are highest. However, our analysis suggests that even under these more optimal conditions, NT yields might be similar to but not exceed CT yields. That said, given other benefits of NT, it may be economical under certain conditions. We conducted the most comprehensive and up-to-date global meta-analysis on NT impacts in rice since 2013. This study evaluates how soil properties, agronomic management, and climatic conditions influence yield differences between NT and conventional tillage (CT) and identifies the key conditions under which NT may be agronomically viable. • Global analysis of 579 paired observations (NT vs. CT) from 115 studies (1985 – 2025). • No-till caused a 4.8% decrease in yields compared to conventional tillage. • No-till yields improved in soils with fine texture, high carbon, and neutral soil pH. • No-till systems improved at higher N rates but were still below CT yields. • Rice establishment system, rotation, residues, and climate zone showed no effects.

Context and objective: To facilitate sustained productivity gains in wheat and maize across Europe, we analysed a long-term (2003–2018) European Value for Cultivation and Use (VCU) dataset to quantify genotypic (G) and genotype-by-environment interaction (GEI) variation and to assess variety-testing precision for grain yield. Methods: Mixed-model analyses were applied to partition genetic and environmental sources of yield variation across European VCU networks in seven countries and to estimate variety-comparison precision (LSDs) and annual genetic trends for both crops. Results: Across Europe, G and GEI variances were comparable, explaining in total 10–29 % of total phenotypic variance in wheat and 8–29 % in maize. LSDs ranged 0.29–0.56 t ha⁻¹ and 0.38–0.81 t ha⁻¹, respectively. Genetic trends ranged 0.45–1.81 % yr⁻¹ for wheat and 0.90–1.31 % yr⁻¹ for maize. Testing precision allowed effective comparison of wheat and maize varieties differing by 2.7–10.5 and 3.3–6.4 release years, respectively, aligning with typical breeding turnover rates. Stronger GEI was associated with larger genetic trends in maize yield, whereas in wheat, observed trends were substantially smaller than the maximum potential implied by GEI magnitudes. Combining data from national VCU networks improved precision and allowed for efficient comparison of varieties not jointly tested. Conclusions: European VCU networks are well-optimized for the number of testing locations given national evaluation periods (2–3 years). Strengthening connections among national VCU networks could further improve precision, while integrating environmental and genetic information would enable a predictive testing system, accelerating breeding progress and the development of resilient varieties. Significance: This first pan-European appraisal of current VCU testing efficiency for two major crops provides a quantitative basis for optimizing variety evaluation across Europe.