Machine learning unveils the role of biochar application in enhancing tea yield by mitigating soil acidification in tea plantations.
Journal:
The Science of the total environment
PMID:
39884194
Abstract
Biochar, a widely utilized soil amendment in environmental applications, has been employed to enhance tea cultivation. This study utilized three machine learning models to investigate the effects of biochar on tea growth and yield, with the random forest (RF) model demonstrating superior performance (R = 0.8768, Root Mean Square Error = 6.1537). Feature importance analysis revealed that biochar characteristics and experimental conditions constitute critical factors exerting an impact on the output, accounting for 39.2 % and 38.6 %, respectively. Specifically, the Ca content of biochar (weight 0.274), the quantity of biochar applied (weight 0.206), and the calcium (Ca) content of soil (weight 0.120) emerged as the three most significant factors affecting tea yield. In conclusion, the machine learning models developed in this study elucidate the multifactorial impact of biochar application on tea yield, providing theoretical and methodological support for practical biochar application strategies in tea production.