IntroductionPromoting sustainable crop production is enhanced by an effective method to assess soil health. However, soil health assessment is challenging due to multiple interactions among dynamic soil properties (i.e., soil health indicators) across management practices and agroecological regions. We tested several currently popular soil health assessment methods for cropping systems in Tennessee in the southeastern US and found that these methods failed to differentiate Tennessee soil health under long-term conservation and conventional management.Materials and methodsThis study developed a Tennessee weighted soil health index (WSHI) by: 1) selecting a set of management-sensitive soil health indicators, 2) assigning meaningful weights to indicators, and 3) normalizing the scores based on regionally relevant undisturbed natural reference sites. The tested cropping systems treatments were moldboard plow (MP) in continuous soybean (SS), no tillage (NT) in SS, NT with wheat cover (NTW) in SS, no cover and chisel plow (NCCT) in continuous cotton (CC), no cover and no tillage (NCNT) in CC, and hairy vetch cover and no tillage (VCNT) in CC. In addition, two woodlots and one grassland sites in the vicinity of the cropping systems were selected to represent undisturbed natural systems.Results and discussionOut of 22 indicators that proved to be management-sensitive, six were selected as a minimum dataset (MDS). These were particulate organic matter C (POM-C), soil respiration from 4-day incubation (4d CO2), small macroaggregate (0.250-2mm)-associated C (SMA-C), surface hardness (PR15), microbial biomass N (MBN), and bulk density (BD). Measured values of the MDS indicators were transformed into unitless normalized scores (based on the regional range of the indicator), and finally integrated into WSHI scores using a weighted-addition approach. Additionally, the soil health gap (SHG) between the soil health of the regional reference system and different cropping systems was calculated. Results revealed that WSHI strongly differentiated soil health between long-term conservation and conventional managements practices. The WSHI scores for southeastern cropland soils varied as follows: VCNT = NTW > NT > NCNT ≥ NCCT ≥ MP. The SHGs under MP, NCCT, NCNT, NT, NTW, and VCNT were 85.5, 79.9, 68, 45.1, 25.2, and 24.3, respectively, relative to the average WSHI of three undisturbed systems. Results showed that the WSHI approach is effective in more meaningful regional assessment of soil health and SHG can be a potential metric for comparing soil health across agroecological regions.