Scholarlink

<p>The improvement of strategies for estimating the nutrient composition of arable crops is currently a fundamental prerequisite for sensible and consistent management of soil properties. Modeling methods can be used to assess the soil properties of arable land and to investigate crop properties via soil operations. Therefore, the main objective of this experiment was to investigate the use of Support Vector Regression to predict oil content and fatty acids in purslane seeds and the uptake of elements in response to soil macronutrient content under the influence of fertilizer types. Accordingly, two field trials were conducted with purslane using chemical and nanoparticulate fertilizers (nitrogen, phosphorus and potassium) and vermicompost (0, 5, 10 and 15&#xa0;t&#xa0;ha<sup>−1</sup>). The results showed that the accuracy of the prediction model in response to soil potassium was more than 95% (R<sup>2</sup> = 0.957) when the oleic acid datasets were used as input. Moreover, the best soil potassium content was between 160 and 300&#xa0;mg&#xa0;kg<sup>−1</sup>. Similarly, the best oleic acid content was between 22.06 and 24.89% in response to soil potassium obtained with chemical fertilizers and NPK nanoparticles and 5 and 15&#xa0;t&#xa0;ha<sup>−1</sup> of vermicompost. The importance of oleic acid in response to soil potassium is also identified as a factor that significantly influences purslane prediction models. Higher levels of soil nutrients do not contribute to increased oil content and fatty acids in the seeds, and over-application may not be economical.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Nano and chemical fertilizer in conjunction with vermicompost as an effective organic fertilizer using support vector regression: uptake of nutrient elements, and enhancing oil content and fatty acids in Portulaca oleracea

  • 期刊论文

  • 发表日期：

    2025年10月8日

查看全文
  • Mohsen Noori,
  • Salim Farzaneh,
  • Raouf Seyed Sharifi,
  • Abbas Keshtehgar,
  • Farnaz Ahmadi-Nouraldinvand,
  • Alireza Shahriari

摘要

The improvement of strategies for estimating the nutrient composition of arable crops is currently a fundamental prerequisite for sensible and consistent management of soil properties. Modeling methods can be used to assess the soil properties of arable land and to investigate crop properties via soil operations. Therefore, the main objective of this experiment was to investigate the use of Support Vector Regression to predict oil content and fatty acids in purslane seeds and the uptake of elements in response to soil macronutrient content under the influence of fertilizer types. Accordingly, two field trials were conducted with purslane using chemical and nanoparticulate fertilizers (nitrogen, phosphorus and potassium) and vermicompost (0, 5, 10 and 15 t ha−1). The results showed that the accuracy of the prediction model in response to soil potassium was more than 95% (R2 = 0.957) when the oleic acid datasets were used as input. Moreover, the best soil potassium content was between 160 and 300 mg kg−1. Similarly, the best oleic acid content was between 22.06 and 24.89% in response to soil potassium obtained with chemical fertilizers and NPK nanoparticles and 5 and 15 t ha−1 of vermicompost. The importance of oleic acid in response to soil potassium is also identified as a factor that significantly influences purslane prediction models. Higher levels of soil nutrients do not contribute to increased oil content and fatty acids in the seeds, and over-application may not be economical.