Phenological Stages Analysis in Grapevines Using an Electronic Nose

IF 1.4 Q3 AGRONOMY
Alan Fernando Coelho Garcia, Ricardo Antonio Ayub, José Carlos Ferreira Da Rocha, Hugo Valadares Siqueira, Sergio Luiz Stevan
{"title":"Phenological Stages Analysis in Grapevines Using an Electronic Nose","authors":"Alan Fernando Coelho Garcia,&nbsp;Ricardo Antonio Ayub,&nbsp;José Carlos Ferreira Da Rocha,&nbsp;Hugo Valadares Siqueira,&nbsp;Sergio Luiz Stevan","doi":"10.1007/s40003-024-00730-w","DOIUrl":null,"url":null,"abstract":"<div><p>The vineyards present different phenological phases that comprise dormancy, bud break, and flowering buds going through different stages of development, such as inflorescence formation, flowering, fruit set, growth and fruit maturation. To control the quantity and quality of production, thinning is used in table grapes. The technique reduces berry number to improve fruit growth, but it is costly and in some cases impractical in the entire extension of an orchard. The right moment for execution and the intensity are complex issues that involve specific knowledge about the conditions of the vineyard. Therefore, phenological information that can help planning and decision-making about thinning is relevant and can improve the cost-effectiveness of the technique in viticulture. An electronic nose system was developed to collect and analyze compound volatile variations during the growing season, more specifically during the period of bud growth and ripening in three grape cultivars (BRS Vitória, Niagara Rosada, Bordô). The data were collected from October 2021 to February 2022. The research hypothesis is that the electronic nose can identify the general stage of plant development. To verify the hypothesis, a classification analysis was performed for each cultivar. The result showed that all models presented balanced accuracy above 85% for the cultivar BRS Vitória, above 92% for Niagara, and above 93% for Bordô, with better performance for models based on <i>K-nearest neighbors </i> (KNN), and <i>random forest</i>, than those based on <i>extreme learning machine</i> and <i>support vector machine</i>. In the total of 24 models, 9 for BRS Vitória, 9 for Niagara, and 11 for Bordô did not obtain error given the metrics used. It was observed that the normalization of the database is not necessary to improve the accuracy rates obtained, which obtained total rates using the KNN classifier. Regarding the research hypothesis, it is considered that the electronic nose is capable of distinguishing between the different stages proposed for each analyzed cultivar and between them. The results of this work indicate a potential use of the electronic nose to aid decision-making in vineyard activities.</p></div>","PeriodicalId":7553,"journal":{"name":"Agricultural Research","volume":"13 4","pages":"636 - 653"},"PeriodicalIF":1.4000,"publicationDate":"2024-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Agricultural Research","FirstCategoryId":"1085","ListUrlMain":"https://link.springer.com/article/10.1007/s40003-024-00730-w","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"AGRONOMY","Score":null,"Total":0}
引用次数: 0

Abstract

The vineyards present different phenological phases that comprise dormancy, bud break, and flowering buds going through different stages of development, such as inflorescence formation, flowering, fruit set, growth and fruit maturation. To control the quantity and quality of production, thinning is used in table grapes. The technique reduces berry number to improve fruit growth, but it is costly and in some cases impractical in the entire extension of an orchard. The right moment for execution and the intensity are complex issues that involve specific knowledge about the conditions of the vineyard. Therefore, phenological information that can help planning and decision-making about thinning is relevant and can improve the cost-effectiveness of the technique in viticulture. An electronic nose system was developed to collect and analyze compound volatile variations during the growing season, more specifically during the period of bud growth and ripening in three grape cultivars (BRS Vitória, Niagara Rosada, Bordô). The data were collected from October 2021 to February 2022. The research hypothesis is that the electronic nose can identify the general stage of plant development. To verify the hypothesis, a classification analysis was performed for each cultivar. The result showed that all models presented balanced accuracy above 85% for the cultivar BRS Vitória, above 92% for Niagara, and above 93% for Bordô, with better performance for models based on K-nearest neighbors (KNN), and random forest, than those based on extreme learning machine and support vector machine. In the total of 24 models, 9 for BRS Vitória, 9 for Niagara, and 11 for Bordô did not obtain error given the metrics used. It was observed that the normalization of the database is not necessary to improve the accuracy rates obtained, which obtained total rates using the KNN classifier. Regarding the research hypothesis, it is considered that the electronic nose is capable of distinguishing between the different stages proposed for each analyzed cultivar and between them. The results of this work indicate a potential use of the electronic nose to aid decision-making in vineyard activities.

利用电子鼻分析葡萄树的物候期
葡萄园有不同的物候期,包括休眠期、花芽分化期和花蕾发育期,如花序形成期、开花期、坐果期、生长期和果实成熟期。为了控制产量和质量,鲜食葡萄采用疏果技术。这项技术可以减少浆果数量,改善果实生长,但成本较高,而且在某些情况下对整个果园来说并不可行。正确的实施时机和强度是一个复杂的问题,涉及到对葡萄园条件的具体了解。因此,能够帮助规划和决策疏果的物候信息非常重要,可以提高葡萄栽培技术的成本效益。我们开发了一套电子鼻系统,用于收集和分析三个葡萄栽培品种(BRS Vitória、Niagara Rosada 和 Bordô)在生长季节的化合物挥发变化,特别是芽生长和成熟期的化合物挥发变化。数据收集时间为 2021 年 10 月至 2022 年 2 月。研究假设是,电子鼻可以识别植物发育的一般阶段。为了验证这一假设,对每个栽培品种进行了分类分析。结果表明,所有模型的准确率均衡,BRS Vitória的准确率超过85%,Niagara的准确率超过92%,Bordô的准确率超过93%,基于K-近邻(KNN)和随机森林的模型比基于极端学习机和支持向量机的模型表现更好。在总共 24 个模型中,有 9 个维托里亚 BRS 模型、9 个尼亚加拉 BRS 模型和 11 个波尔多 BRS 模型在所使用的指标下没有出现误差。据观察,使用 KNN 分类器获得的总准确率并不需要对数据库进行规范化处理来提高准确率。关于研究假设,我们认为电子鼻能够区分每个被分析栽培品种的不同阶段以及它们之间的不同阶段。这项工作的结果表明,电子鼻可用于葡萄园活动的辅助决策。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
CiteScore
3.80
自引率
0.00%
发文量
24
期刊介绍: The main objective of this initiative is to promote agricultural research and development. The journal will publish high quality original research papers and critical reviews on emerging fields and concepts for providing future directions. The publications will include both applied and basic research covering the following disciplines of agricultural sciences: Genetic resources, genetics and breeding, biotechnology, physiology, biochemistry, management of biotic and abiotic stresses, and nutrition of field crops, horticultural crops, livestock and fishes; agricultural meteorology, environmental sciences, forestry and agro forestry, agronomy, soils and soil management, microbiology, water management, agricultural engineering and technology, agricultural policy, agricultural economics, food nutrition, agricultural statistics, and extension research; impact of climate change and the emerging technologies on agriculture, and the role of agricultural research and innovation for development.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信