Soil fertility assessment for mango production in karst agroecosystems of Southwest China

Accurate assessment of soil fertility is crucial for enhancing mango productivity and quality. However, rare soil fertility surveys have been reported on mango orchards, especially in karst mountainous areas of China. In this study, an investigation was conducted using 152 paired soil-fruit samples collected from commercial mango orchards in Baise City, Guangxi Province, southwestern China. Principal component analysis (PCA) was employed to establish a minimum data set (MDS), and two distinct weighting methods were compared to calculate the soil integrated fertility index (IFI). The 5 representative indicators—pH, available potassium (AK), organic matter (OM), available phosphorus (AP), and available boron (AB)—were selected to construct the MDS. Together with a correlation-based weighting method, the average IFI was determined to be 0.51 (range: 0.17–0.92), grading the overall fertility as “Medium.” Significant positive correlations were observed between IFI and both yield (R 2 = 0.45, p < 0.01) and fruit quality index (FQI) (R 2 = 0.49, p < 0.01). Notably, AK and OM were identified as the primary determinants of yield and FQI using random forest model, with 20 g/kg for OM or 120 mg/kg for AK as the critical thresholds for achieving high yields. Furthermore, AP, AB, and altitude were determined as the secondary limiting factors, with corresponding high yield thresholds of 20 mg/kg, 0.30 mg/kg, and 300 m, respectively. These findings provide a theoretical basis and practical guidelines for optimizing orchard establishment and nutrient management in karst agroecosystems. • Five factors (pH, AK, OM, AP, and AB) were selected to establish MDS. • Correlation-based weighting outperformed communality-based method. • IFI correlated positively with mango yield and quality. • Critical thresholds for high yield were 20 g/kg of OM and 120 mg/kg of AK.