Scholarlink

<p>Verification of geographical origin and identification of region-specific properties of honey have gained increasing attention due to their influence on consumer preferences and market value. In this study, it was aimed to classify a total of 121 unifloral, multifloral, and honeydew honeys collected from different geographical regions of Türkiye using a linear discriminant analysis (LDA) model. The stepwise function was employed to identify the key variables based on physicochemical, chemical, and biological properties contributing most significantly to geographical differentiation. Except for titratable acidity and proline content, the examined honey properties, including basic physicochemical properties, amino acid profiles, and biological activities, showed significant differences among the seven geographical regions. The honey samples exhibited wide botanical diversity, and their quality parameters were largely compliant with legal standards. Honey samples collected from Black Sea region showed higher total phenolic and flavonoid contents and antioxidant activity. Aegean honeys exhibited stronger anti-hyaluronidase effects, and Eastern Anatolia honeys demonstrated greater inhibitory effect against <i>Staphylococcus aureus</i> ATCC 25923. The LDA model achieved classification accuracies of 91.7% for original cases and 85.5% after cross-validation. Function 1 was mainly associated with higher aspartic acid content and lower water activity, while Function 2 was characterized by stronger <i>S. aureus</i> ATCC 25923 inhibition, higher cysteine, and lower aspartic acid levels. The honey samples collected from Marmara (94.1%) and Central Anatolia (95.2%) were clearly distinguishable, while those from the Mediterranean (73.5%) and Black Sea (80.0%) regions showed lower cross-validated discrimination, particularly in relation to the Central Anatolia samples, due to overlapping properties.</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

Unveiling the geographical signatures of Turkish honey through multivariate analysis of its physicochemical, chemical, and biological properties

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月28日

查看全文
  • Ceren Mutlu,
  • Cihadiye Candal-Uslu,
  • Andaç Koç,
  • Sultan Arslan-Tontul,
  • Duygu Gizem Bilgin-Çavana,
  • Meltem Öztürk,
  • Fatma Ersöz,
  • Aslı Özkök,
  • Mustafa Erbaş

摘要

Verification of geographical origin and identification of region-specific properties of honey have gained increasing attention due to their influence on consumer preferences and market value. In this study, it was aimed to classify a total of 121 unifloral, multifloral, and honeydew honeys collected from different geographical regions of Türkiye using a linear discriminant analysis (LDA) model. The stepwise function was employed to identify the key variables based on physicochemical, chemical, and biological properties contributing most significantly to geographical differentiation. Except for titratable acidity and proline content, the examined honey properties, including basic physicochemical properties, amino acid profiles, and biological activities, showed significant differences among the seven geographical regions. The honey samples exhibited wide botanical diversity, and their quality parameters were largely compliant with legal standards. Honey samples collected from Black Sea region showed higher total phenolic and flavonoid contents and antioxidant activity. Aegean honeys exhibited stronger anti-hyaluronidase effects, and Eastern Anatolia honeys demonstrated greater inhibitory effect against Staphylococcus aureus ATCC 25923. The LDA model achieved classification accuracies of 91.7% for original cases and 85.5% after cross-validation. Function 1 was mainly associated with higher aspartic acid content and lower water activity, while Function 2 was characterized by stronger S. aureus ATCC 25923 inhibition, higher cysteine, and lower aspartic acid levels. The honey samples collected from Marmara (94.1%) and Central Anatolia (95.2%) were clearly distinguishable, while those from the Mediterranean (73.5%) and Black Sea (80.0%) regions showed lower cross-validated discrimination, particularly in relation to the Central Anatolia samples, due to overlapping properties.