Establishing and verifying the nutrient expert system for citrus fertilization recommendation

Fertilization is one of the most important aspects during commercial citrus production. A better tradeoff is also required for the fruit quality with the fertilizer reduction and nutrients efficiency. In this study, an expert system was constructed for the citrus fertilization recommendation of mineral nutrients, according to the comprehensive evaluation of fruit quality. The 40 plots of 10-12a Longan pomelo (Citrus Grandis cv. 'Longan') orchard were selected in the Modern Agricultural Industrial Park of Longan Pomelo, Guangan City, Sichuan Province, China. The indexes were tracked from 2019 to 2020, including the tree mineral nutrition at the fruit expansion stage (early June), and the fruit quality at the fruit maturity stage (mid-November). The original database was then established. Three algorithms were constructed, including the nutrient criteria generation of high-quality leaf at the fruit expansion stage, the correction fertilization of deviation from the optimum percentage (DOP), according to the standard of "yield quantification, the total quantity control, and staging control", and the compound fertilizer formula recommendation using the local fertilizer and agricultural materials supply database. Four functions were then achieved, namely, the automatic generation of high-quality leaf nutrient criteria during fruit expansion, the nutritional assessment of citrus orchards, the recommendation of available nutrient formulations, and the recommendation of compound fertilizer formulations. The system was distributed to the growers, government agencies, and system administrators with the different permissions using role-based access control (RBAC). A series of validation experiments were carried out on the nutrient expert system for the citrus fertilization recommendation at the early stage of fruit expansion in 2021. Two circus orchards were selected, where the chemical fertilizer was applied in the Longan Town, whereas, the organic fertilizer was in the Chongwang Town. The average low-, middle-, and high-yield plantations were taken as the system target yields. The effective formula of nutrient correction was used to generate three schemes of recommended fertilization for two orchards, labeled as the RF1, RF2, RF3. The FP control groups were marked as the experience fertilization schemes for the citrus orchards in Longan and Chongwang Town, respectively. The results showed that the RF2 treatments for two orchards performed the best with the average yield of 35.71 t/hm2, particularly serving as the recommended fertilization scheme. Specifically, the RF2 treatment for Longan was reduced by 40.3%, 43.6%, and 1.3% in the nitrogen (N), Kalium (K), and potassium (P) inputs, respectively, compared with the chemical experience fertilization schemes. By contrast, the RF2 treatment for Chongwang was reduced the N, K and P inputs by 27.1%, 30.8%, and 78.8%. Those two RF2 treatments were achieved in the optimal purpose of "Fertilizer reduction". At the same time, the treatments can be expected to effectively avoid the delays in the fruit coloring and the low overall fruit quality that caused by the excessive and imbalanced chemical fertilization inputs. The low inputs of N and P fertilizers can be alleviated the small fruit sizes and delayed fruit pulp color, due to the empirical organic fertilizer application. Therefore, the nutrient fertilization recommendation system can be promising potential to promote the nutrient correction formula fertilization for commercial citrus industry.