Evaluation of grafting compatibility and prediction modelling in Sapindus

Abstract Grafting compatibility refers to the successful healing between stock and scion, as well as the normal functioning of their tissues. This article investigates the survival rates and growth conditions of various Sapindus grafting combinations and establishes a new evaluation system for Sapindus grafting compatibility. This system aims to provide theoretical support for predicting Sapindus grafting compatibility and enhancing the development of improved varieties. The study utilizes 15 Sapindus grafting combinations as research materials, assesses their survival rates and growth conditions, employs the CRITIC method to calculate weights, and develops a Sapindus grafting compatibility evaluation model. By analyzing the physiological characteristics of the scion leaves from these 15 grafting combinations, variance analysis, cluster analysis, and linear regression models were utilized to assess the accuracy of the grafting compatibility evaluation model. The results indicate that the growth of grafted plants serves as a critical evaluation index of grafting compatibility, providing a more scientific approach than relying solely on survival rate metrics. This research offers a theoretical foundation and data support for predicting Sapindus grafting compatibility; an effective prediction model will substantially enhance the efficiency of breeding and promoting improved varieties, thereby fostering the growth of the Sapindus industry.