Skip to main content
Dryad

Data and code from: Soybean aphid (Hemiptera: Aphididae) population dynamics are associated with the temporal scale and pattern of soil moisture in soybean fields

Yang, Dan-Hong1; Chang, Jui-Tse1; Yang, Kai-Wen1; Su, Tzu-Jiun1; Wang, Wei-Chou1; Tseng, Yu-Hsin2; Huang, Meng-Yuan2; Tan, Ching-Wen2; Wu, Steven H.1; Lin, Po-An1

Research facility: Beigou Agricultural Experiment Station of National Chung Hsing University (Wufeng District, Taichung, Taiwan)
Published Jul 28, 2026 on Dryad. https://doi.org/10.5061/dryad.0k6djhbgw

Data files

Jul 28, 2026 version files 20.55 MB

Click names to download individual files

Abstract

Understanding and predicting pest outbreaks is central to sustainable crop management, yet field evidence linking belowground abiotic conditions to aboveground herbivore dynamics remains limited, particularly across the temporal scales relevant to plant physiological responses. This dataset provides high-temporal-resolution soil environmental measurements and field observations of soybean aphid (Aphis glycines Matsumura) abundance collected in a soybean field in Taiwan over two growing seasons. Using an Internet of Things sensor network, soil moisture and soil electrical conductivity (EC) were monitored continuously, while soybean aphid abundance was sampled weekly. The dataset was used to evaluate associations between soybean aphid abundance and soil conditions summarized across multiple temporal windows, ranging from same-day values to 7-d rolling averages. R scripts are provided for data processing, statistical analyses, model evaluation, and figure generation. After accounting for seasonal progression and plant ontogeny, soil moisture showed more consistent, although generally weak, associations with soybean aphid abundance than soil EC. Short-term moisture metrics, particularly daily to 3-d averages, showed the most consistent positive associations, whereas associations over longer integration periods were weaker and more season-dependent. Predictive analyses showed that moisture-based models generally performed better in cross-validation than EC-only models, while adding EC provided limited improvement beyond soil moisture alone. Because the study was observational and conducted at a single field site, these results represent conditional field associations rather than direct evidence of causation. This dataset and accompanying R code provide a reproducible resource for investigating how the temporal scale of soil environmental measurements influences inference about soybean aphid–soil associations and for evaluating the use of high-frequency soil sensing in agricultural pest monitoring.