Soil enzyme activities are known for early indicators of soil health but have not been related to most soil properties and crop yields. We examined the relationships among activities of four soil enzymes (β‐glucosidase [BG], N‐acetyl‐β‐glucosaminidase [NAG], arylsulfatase [AST], and phosphomonoesterase [PME]), 62 soil properties, and mean crop yields in two long‐term (14‐ and 36‐yr‐old) dryland farming sites in the northern Great Plains, USA. Treatments were no‐till and till crop rotations with spring wheat (Triticum aestivum L.), barley (Hordeum vulgare L.), pea (Pisum sativum L.), and fallow with and without N fertilization. Soil samples collected before farming operations in April 2019 were analyzed for physical, chemical, biological, and biochemical properties and mean crop yields were determined. The principal component analysis (PCA) showed that BG was associated to most soil physical, chemical, biological, and biochemical properties followed by NAG and PME at both sites. The AST was associated to limited number of soil properties. The BG related to mean crop yield across years in one site and NAG related to yields in both sites, but AS and PME were not related to yields. Because of the greater sensitivity to cropping systems and relationships with most soil properties and long‐term crop yields, BG and NAG activities may be considered as promising soil health indicators compared with other enzyme activities under dryland cropping systems in the northern Great Plains.

Carbon dioxide flush after rewetting of dried soils has been recommended as a promising soil health indicator, but it has not been related to most soil properties and crop yields. We evaluated the effect of cropping systems and N fertilization on CO 2 flushes at 1-(1dC) and 4-d incubations (4dC) after rewetting of dried soils and related to 54 soil physical, chemical, and biological properties and annualized crop yields in two long-term experimental sites in eastern Montana (USA). Treatments included till and no-till spring wheat (Triticum aestiveum L.), pea (Pisum sativum L.), and fallow rotations with and without N fertilization. Carbon dioxide flushes were lower in till crop-fallow than in no-till continuous cropping systems at both sites. The 1dC was correlated to 5 soil physical, 7 chemical, and 12 biological properties, and 4dC was correlated to 9 physical, 8 chemical, and 11 biological properties in Froid. In Sidney, 1dC was correlated to 10 physical, 13 chemical, and 9 biological properties, and 4dC was correlated to 7 physical, 11 chemical, and 2 biological properties (1-8 moderately, 18-21 strongly, and 1-3 very strongly related). Carbon dioxide flushes were also related to mean annualized crop yields in both sites, except for the relationship between 4dC and crop yield in Sidney. Because of its stronger relationship with soil properties and crop yields, 1dC after rewetting of dried soils determined by using the infrared gas analyzer can be used as a simple, rapid, reliable, and inexpensive indicator of measuring soil health in dryland cropping systems. INTRODUCTIONCarbon dioxide flush after rewetting of dried soil, which measures microbial activity, has been proposed as one of the Abbreviations: 1dC, CO 2 flush at 1-d incubation determined by using the infrared analyzer; 4dC, CO 2 flush at 4-d incubation determined by the alkali-trap method; ACEP, autoclaved citrate-extractable protein;