Applying deep neural networks to predict incidence and phenology of plant pests and diseases
Abstract A major challenge of agriculture is to improve the sustainability of food production systems in order to provide enough food for a growing human population. Pests and pathogens cause vast yield losses, while crop protection practices raise environmental and human health concerns. Decision s...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2021-10-01
|
| Series: | Ecosphere |
| Subjects: | |
| Online Access: | https://doi.org/10.1002/ecs2.3791 |
| _version_ | 1818741913841303552 |
|---|---|
| author | Marc Grünig Elisabeth Razavi Pierluigi Calanca Dominique Mazzi Jan Dirk Wegner Loïc Pellissier |
| author_facet | Marc Grünig Elisabeth Razavi Pierluigi Calanca Dominique Mazzi Jan Dirk Wegner Loïc Pellissier |
| author_sort | Marc Grünig |
| collection | DOAJ |
| description | Abstract A major challenge of agriculture is to improve the sustainability of food production systems in order to provide enough food for a growing human population. Pests and pathogens cause vast yield losses, while crop protection practices raise environmental and human health concerns. Decision support systems provide detailed information on optimal timing and necessity of crop protection interventions, but are often based on phenology models that are time‐, cost‐, and labor‐intensive in development. Here, we aim to develop a data‐driven approach for pest damage forecasting, relying on big data and deep learning algorithms. We present a framework for the development of deep neural networks for pest and pathogen damage classification and show their potential for predicting the phenology of damages. As a case study, we investigate the phenology of the pear leaf blister moth (Leucoptera malifoliella, Costa). We employ a set of 52,322 pictures taken during a period of 19 weeks and establish deep neural networks to categorize the images into six main damage classes. Classification tools achieved good performance scores overall, with differences between the classes indicating that the performance of deep neural networks depends on the similarity to other damages and the number of training images. The reconstructed damage phenology of the pear leaf blister moth matches mine counts in the field. We further develop statistical models to reconstruct the phenology of damages with meteorological data and find good agreement with degree‐day models. Hence, our study indicates a yet underexploited potential for data‐driven approaches to enhance the versatility and cost efficiency of plant pest and disease forecasting. |
| first_indexed | 2024-12-18T02:04:11Z |
| format | Article |
| id | doaj.art-bcbf6a1e174c4f58b20746466466418a |
| institution | Directory Open Access Journal |
| issn | 2150-8925 |
| language | English |
| last_indexed | 2024-12-18T02:04:11Z |
| publishDate | 2021-10-01 |
| publisher | Wiley |
| record_format | Article |
| series | Ecosphere |
| spelling | doaj.art-bcbf6a1e174c4f58b20746466466418a2022-12-21T21:24:40ZengWileyEcosphere2150-89252021-10-011210n/an/a10.1002/ecs2.3791Applying deep neural networks to predict incidence and phenology of plant pests and diseasesMarc Grünig0Elisabeth Razavi1Pierluigi Calanca2Dominique Mazzi3Jan Dirk Wegner4Loïc Pellissier5Agroscope RD Plant Protection Wädenswil SwitzerlandAgroscope RD Plant Protection Wädenswil SwitzerlandAgroscope RD Agroecology and Environment Zurich SwitzerlandAgroscope RD Plant Protection Wädenswil SwitzerlandEcoVision Lab ETH Zurich Zurich SwitzerlandLandscape Ecology ETH Zurich Zurich SwitzerlandAbstract A major challenge of agriculture is to improve the sustainability of food production systems in order to provide enough food for a growing human population. Pests and pathogens cause vast yield losses, while crop protection practices raise environmental and human health concerns. Decision support systems provide detailed information on optimal timing and necessity of crop protection interventions, but are often based on phenology models that are time‐, cost‐, and labor‐intensive in development. Here, we aim to develop a data‐driven approach for pest damage forecasting, relying on big data and deep learning algorithms. We present a framework for the development of deep neural networks for pest and pathogen damage classification and show their potential for predicting the phenology of damages. As a case study, we investigate the phenology of the pear leaf blister moth (Leucoptera malifoliella, Costa). We employ a set of 52,322 pictures taken during a period of 19 weeks and establish deep neural networks to categorize the images into six main damage classes. Classification tools achieved good performance scores overall, with differences between the classes indicating that the performance of deep neural networks depends on the similarity to other damages and the number of training images. The reconstructed damage phenology of the pear leaf blister moth matches mine counts in the field. We further develop statistical models to reconstruct the phenology of damages with meteorological data and find good agreement with degree‐day models. Hence, our study indicates a yet underexploited potential for data‐driven approaches to enhance the versatility and cost efficiency of plant pest and disease forecasting.https://doi.org/10.1002/ecs2.3791decision support systemdeep neural networkimage classificationinsect pestphenological modeling |
| spellingShingle | Marc Grünig Elisabeth Razavi Pierluigi Calanca Dominique Mazzi Jan Dirk Wegner Loïc Pellissier Applying deep neural networks to predict incidence and phenology of plant pests and diseases Ecosphere decision support system deep neural network image classification insect pest phenological modeling |
| title | Applying deep neural networks to predict incidence and phenology of plant pests and diseases |
| title_full | Applying deep neural networks to predict incidence and phenology of plant pests and diseases |
| title_fullStr | Applying deep neural networks to predict incidence and phenology of plant pests and diseases |
| title_full_unstemmed | Applying deep neural networks to predict incidence and phenology of plant pests and diseases |
| title_short | Applying deep neural networks to predict incidence and phenology of plant pests and diseases |
| title_sort | applying deep neural networks to predict incidence and phenology of plant pests and diseases |
| topic | decision support system deep neural network image classification insect pest phenological modeling |
| url | https://doi.org/10.1002/ecs2.3791 |
| work_keys_str_mv | AT marcgrunig applyingdeepneuralnetworkstopredictincidenceandphenologyofplantpestsanddiseases AT elisabethrazavi applyingdeepneuralnetworkstopredictincidenceandphenologyofplantpestsanddiseases AT pierluigicalanca applyingdeepneuralnetworkstopredictincidenceandphenologyofplantpestsanddiseases AT dominiquemazzi applyingdeepneuralnetworkstopredictincidenceandphenologyofplantpestsanddiseases AT jandirkwegner applyingdeepneuralnetworkstopredictincidenceandphenologyofplantpestsanddiseases AT loicpellissier applyingdeepneuralnetworkstopredictincidenceandphenologyofplantpestsanddiseases |