Effective detection of stored-grain insect pests (SGIP) constitutes a critical component of modern grain storage management, directly impacting food security and minimizing post-harvest losses. However, achieving accurate detection through object detection methods remains challenging due to small target scales, subtle inter-species morphological differences, and complex background interference. Therefore, this study proposes a high-precision SGIP model based on Transformer architecture called Multi-Species Stored-Grain Insect Pest DEtection TRansformer (MSGP-DETR). To enhance small target detection capabilities, the network incorporates a Discrete Cosine Transform Small Target Enhancement Module (DSTEM) for frequency-domain feature enhancement. Additionally, a Multi-Point Spiral Mamba (MPSM) module is designed to capture long-range dependencies through linear computational complexity. Moreover, an Adaptive Partial Enhancement Lightweight Aggregation Network (APELAN) is developed to reduce computational complexity while preserving essential feature information. To evaluate MSGP-DETR's performance in SGIP detection, we collected adult live samples of eight common SGIP species and constructed a dataset named SGIP8. Experimental results demonstrate that MSGP-DETR achieves 94.6 % AP50 and 50.1 % AP50:95 on the SGIP8 dataset, with only 9.2M parameters and 19.7 GFLOPs computational complexity. The detection speed reaches 36.6 FPS, with overall performance surpassing current mainstream models. Finally, validation on additional Pest24 and VisDrone2019 datasets confirms the effectiveness of MSGP-DETR. This research provides technical support for automated grain storage management.
{"title":"MSGP-DETR: A multi-species stored-grain insect pest detection transformer integrating frequency-domain enhancement and state space models","authors":"Zhizhong Guan , Fuyan Sun , Zongwang Lyu , Yiyang Xin , Zihan Zhao","doi":"10.1016/j.jspr.2025.102941","DOIUrl":"10.1016/j.jspr.2025.102941","url":null,"abstract":"<div><div>Effective detection of stored-grain insect pests (SGIP) constitutes a critical component of modern grain storage management, directly impacting food security and minimizing post-harvest losses. However, achieving accurate detection through object detection methods remains challenging due to small target scales, subtle inter-species morphological differences, and complex background interference. Therefore, this study proposes a high-precision SGIP model based on Transformer architecture called Multi-Species Stored-Grain Insect Pest DEtection TRansformer (MSGP-DETR). To enhance small target detection capabilities, the network incorporates a Discrete Cosine Transform Small Target Enhancement Module (DSTEM) for frequency-domain feature enhancement. Additionally, a Multi-Point Spiral Mamba (MPSM) module is designed to capture long-range dependencies through linear computational complexity. Moreover, an Adaptive Partial Enhancement Lightweight Aggregation Network (APELAN) is developed to reduce computational complexity while preserving essential feature information. To evaluate MSGP-DETR's performance in SGIP detection, we collected adult live samples of eight common SGIP species and constructed a dataset named SGIP8. Experimental results demonstrate that MSGP-DETR achieves 94.6 % AP<sub>50</sub> and 50.1 % AP<sub>50:95</sub> on the SGIP8 dataset, with only 9.2M parameters and 19.7 GFLOPs computational complexity. The detection speed reaches 36.6 FPS, with overall performance surpassing current mainstream models. Finally, validation on additional Pest24 and VisDrone2019 datasets confirms the effectiveness of MSGP-DETR. This research provides technical support for automated grain storage management.</div></div>","PeriodicalId":17019,"journal":{"name":"Journal of Stored Products Research","volume":"116 ","pages":"Article 102941"},"PeriodicalIF":2.7,"publicationDate":"2025-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145880500","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-27DOI: 10.1016/j.jspr.2025.102944
Sara Ashagheh Bashloo , Masumeh Ziaee , Fariba Sohrabi
The bean weevil, Callosobruchus maculatus (F.) (Coleoptera: Chrysomelidae), is a significant pest affecting the storage and production of legumes, particularly beans. This study evaluates the efficacy of diatomaceous earth (DE) formulations, both alone and in combination with imidacloprid and deltamethrin, utilized at a ratio of 0.05 %, on the mortality and progeny production of C. maculatus. The research further assessed the impact of these formulations on oviposition inhibition, egg hatching rates, and prevention of adult emergence in cowpeas. DE formulations were tested at the doses of 500 and 1000 mg/kg, with six replicates for each treatment. The lowest lethal time (LT50) was recorded for the Dryasil DE formulation combined with imidacloprid at 4.695 days. Mortality rates of C. maculatus adults increased with prolonged exposure and higher doses. The control and talcum powder treatments recorded the highest egg-laying numbers, with 21 and 19.16 eggs, respectively. In contrast, at 1000 mg/kg, the lowest egg counts were observed in the Dryasil + imidacloprid (7.33), Dryasil + deltamethrin (8.66), and imidacloprid (9) treatments. The highest oviposition inhibition rates at this dose were noted for Dryasil + imidacloprid (65.08 %), Dryasil + deltamethrin (58.73 %), and imidacloprid (57.14 %). Additionally, the lowest adult emergence percentage (21.67 %) and the most extended developmental period from egg to adult (21.92 days) occurred in the Dryasil + imidacloprid treatment. Weight loss due to C. maculatus was also evaluated in these treatments, with the lowest percentage of weight loss (5.37 %) in the Dryasil + imidacloprid treatment. These findings suggest that DE formulations, particularly when combined with imidacloprid, are effective in managing C. maculatus infestations in cowpeas, offering a potential strategy for pest control in legume storage.
{"title":"Combined effect of diatomaceous earth and two different insecticides to control Callosobruchus maculatus","authors":"Sara Ashagheh Bashloo , Masumeh Ziaee , Fariba Sohrabi","doi":"10.1016/j.jspr.2025.102944","DOIUrl":"10.1016/j.jspr.2025.102944","url":null,"abstract":"<div><div>The bean weevil, <em>Callosobruchus maculatus</em> (F.) (Coleoptera: Chrysomelidae), is a significant pest affecting the storage and production of legumes, particularly beans. This study evaluates the efficacy of diatomaceous earth (DE) formulations, both alone and in combination with imidacloprid and deltamethrin, utilized at a ratio of 0.05 %, on the mortality and progeny production of <em>C. maculatus.</em> The research further assessed the impact of these formulations on oviposition inhibition, egg hatching rates, and prevention of adult emergence in cowpeas. DE formulations were tested at the doses of 500 and 1000 mg/kg, with six replicates for each treatment. The lowest lethal time (LT<sub>50</sub>) was recorded for the Dryasil DE formulation combined with imidacloprid at 4.695 days. Mortality rates of <em>C. maculatus</em> adults increased with prolonged exposure and higher doses. The control and talcum powder treatments recorded the highest egg-laying numbers, with 21 and 19.16 eggs, respectively. In contrast, at 1000 mg/kg, the lowest egg counts were observed in the Dryasil + imidacloprid (7.33), Dryasil + deltamethrin (8.66), and imidacloprid (9) treatments. The highest oviposition inhibition rates at this dose were noted for Dryasil + imidacloprid (65.08 %), Dryasil + deltamethrin (58.73 %), and imidacloprid (57.14 %). Additionally, the lowest adult emergence percentage (21.67 %) and the most extended developmental period from egg to adult (21.92 days) occurred in the Dryasil + imidacloprid treatment. Weight loss due to <em>C. maculatus</em> was also evaluated in these treatments, with the lowest percentage of weight loss (5.37 %) in the Dryasil + imidacloprid treatment. These findings suggest that DE formulations, particularly when combined with imidacloprid, are effective in managing <em>C. maculatus</em> infestations in cowpeas, offering a potential strategy for pest control in legume storage.</div></div>","PeriodicalId":17019,"journal":{"name":"Journal of Stored Products Research","volume":"116 ","pages":"Article 102944"},"PeriodicalIF":2.7,"publicationDate":"2025-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145836677","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-25DOI: 10.1016/j.jspr.2025.102942
Xia Zhou, Cheng Zhong Li, Yan Sun, Huanxinzh Zhang
During summer storage, a chitosan–neem (Azadirachta indica) leaf extract composite coating was evaluated for preserving adzuki bean (Vigna angularis) seeds. HPLC analysis revealed the presence of azadirachtin (1.25 mg/g), nimbin (0.75 mg/g), nimbidin (0.60 mg/g), quercetin (0.55 mg/g), and rutin (0.50 mg/g) in the aqueous neem extract. Over 90 days, seeds were treated with 1 % chitosan, a composite coating (containing 0.5 % chitosan and 0.25 % neem extract), or left untreated (control). Compared to the control and chitosan-alone treatments, the composite coating significantly (P < 0.05) enhanced seed germination (78.66 %), germination index (88.43), and seedling growth (root length: 7.83 cm; shoot length: 6.34 cm). It also suppressed Fusarium (57.89 %), Aspergillus (61.57 %), and Penicillium (53.32 %) and elevated SOD activity (21.48 U/mg protein) and DPPH radical scavenging activity (75.12 %). Moreover, chitosan–neem treatment reduced stress-related physiological indicators, such as MDA (1.42 nmol/g), NO (2.89 μmol/g), and electrical conductivity (0.69 mS/cm). These results indicate that the chitosan–neem composite coating improves adzuki bean seed germination, antioxidant and antifungal protection, and mitigates storage-induced deterioration.
{"title":"Chitosan–neem (Azadirachta indica) extract coating improves germination, antioxidant defense, and antifungal protection of adzuki bean (Vigna angularis) seeds during summer storage","authors":"Xia Zhou, Cheng Zhong Li, Yan Sun, Huanxinzh Zhang","doi":"10.1016/j.jspr.2025.102942","DOIUrl":"10.1016/j.jspr.2025.102942","url":null,"abstract":"<div><div>During summer storage, a chitosan–neem (<em>Azadirachta indica</em>) leaf extract composite coating was evaluated for preserving adzuki bean (<em>Vigna angularis</em>) seeds. HPLC analysis revealed the presence of azadirachtin (1.25 mg/g), nimbin (0.75 mg/g), nimbidin (0.60 mg/g), quercetin (0.55 mg/g), and rutin (0.50 mg/g) in the aqueous neem extract. Over 90 days, seeds were treated with 1 % chitosan, a composite coating (containing 0.5 % chitosan and 0.25 % neem extract), or left untreated (control). Compared to the control and chitosan-alone treatments, the composite coating significantly (P < 0.05) enhanced seed germination (78.66 %), germination index (88.43), and seedling growth (root length: 7.83 cm; shoot length: 6.34 cm). It also suppressed <em>Fusarium</em> (57.89 %), <em>Aspergillus</em> (61.57 %), and <em>Penicillium</em> (53.32 %) and elevated SOD activity (21.48 U/mg protein) and DPPH radical scavenging activity (75.12 %). Moreover, chitosan–neem treatment reduced stress-related physiological indicators, such as MDA (1.42 nmol/g), NO (2.89 μmol/g), and electrical conductivity (0.69 mS/cm). These results indicate that the chitosan–neem composite coating improves adzuki bean seed germination, antioxidant and antifungal protection, and mitigates storage-induced deterioration.</div></div>","PeriodicalId":17019,"journal":{"name":"Journal of Stored Products Research","volume":"116 ","pages":"Article 102942"},"PeriodicalIF":2.7,"publicationDate":"2025-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145836709","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-25DOI: 10.1016/j.jspr.2025.102943
Luisa Fernanda Zuluaga , Angelica Plata-Rueda , Luis Carlos Martínez
The maize weevil, Sitophilus zeamais, is a globally distributed pest of stored grains. This study characterized the Allium sativum essential oil (EO) using gas chromatography–flame ionization detection (GC–FID) and gas chromatography–mass spectrometry (GC–MS), and evaluated its insecticidal activity against S. zeamais. The effects of the essential oil on weevil mortality, survival, food preference, and repellency were investigated. The major compounds identified were diallyl sulfide, diallyl disulfide, dimethyl trisulfide, allyl methyl disulfide, and diallyl tetrasulfide. In dose–mortality bioassays, the A. sativum EO (LD50 = 0.033 μg insect−1), diallyl disulfide (LD50 = 0.026 μg insect−1), and dimethyl trisulfide (LD50 = 10.385 μg insect−1) were found to be lethal to S. zeamais. The survival rate was 99.9 % in untreated adults but decreased to 50.1 %, 46.7 %, and 45.6 % in weevils treated with LD50 of A. sativum EO, dimethyl trisulfide, and diallyl sulfide, respectively. Furthermore, A. sativum EO and its compounds reduced the weevils’ preference for maize grains and exhibited repellent effects. These results indicate that A. sativum EO and its compounds disrupt various physiological and behavioral functions in S. zeamais. This study opens new perspectives for the control of stored-product pests and represents a preliminary step toward the development of green insecticides.
{"title":"Allium sativum essential oil against Sitophilus zeamais (Coleoptera: Curculionidae): composition, insecticidal activity, and behavioral response","authors":"Luisa Fernanda Zuluaga , Angelica Plata-Rueda , Luis Carlos Martínez","doi":"10.1016/j.jspr.2025.102943","DOIUrl":"10.1016/j.jspr.2025.102943","url":null,"abstract":"<div><div>The maize weevil, <em>Sitophilus zeamais</em>, is a globally distributed pest of stored grains. This study characterized the <em>Allium sativum</em> essential oil (EO) using gas chromatography–flame ionization detection (GC–FID) and gas chromatography–mass spectrometry (GC–MS), and evaluated its insecticidal activity against <em>S. zeamais</em>. The effects of the essential oil on weevil mortality, survival, food preference, and repellency were investigated. The major compounds identified were diallyl sulfide, diallyl disulfide, dimethyl trisulfide, allyl methyl disulfide, and diallyl tetrasulfide. In dose–mortality bioassays, the <em>A. sativum</em> EO (LD<sub>50</sub> = 0.033 μg insect<sup>−1</sup>), diallyl disulfide (LD<sub>50</sub> = 0.026 μg insect<sup>−1</sup>), and dimethyl trisulfide (LD<sub>50</sub> = 10.385 μg insect<sup>−1</sup>) were found to be lethal to <em>S. zeamais</em>. The survival rate was 99.9 % in untreated adults but decreased to 50.1 %, 46.7 %, and 45.6 % in weevils treated with LD<sub>50</sub> of <em>A. sativum</em> EO, dimethyl trisulfide, and diallyl sulfide, respectively. Furthermore, <em>A. sativum</em> EO and its compounds reduced the weevils’ preference for maize grains and exhibited repellent effects. These results indicate that <em>A. sativum</em> EO and its compounds disrupt various physiological and behavioral functions in <em>S. zeamais</em>. This study opens new perspectives for the control of stored-product pests and represents a preliminary step toward the development of green insecticides.</div></div>","PeriodicalId":17019,"journal":{"name":"Journal of Stored Products Research","volume":"116 ","pages":"Article 102943"},"PeriodicalIF":2.7,"publicationDate":"2025-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145836708","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-22DOI: 10.1016/j.jspr.2025.102931
Junaid Zafar , Shuhao Zhang , Zhanpeng Zhu , Ziling Wan , Yanqing Li , Ang Li , Ling Huang , Xuewei Zhang , Chen Peng , Shuai Zhong , Fengliang Jin , Xiaoxia Xu
The tobacco beetle, Lasioderma serricorne, is a globally distributed pest that inflicts major economic losses on stored products due its exceptional dietary adaptability. Understanding host-microbe-diet interactions is crucial for developing sustainable biological control strategies targeting insect-associated microorganisms. Here, we employed comparative metagenomic sequencing to characterize gut microbial communities in L. serricorne larvae reared on four dietary treatments i.e., artificial diet; Artemisia argyi; Nicotiana tabacum and Angelica sinensis. Taxonomic diversity, functional profiles, carbohydrate-active enzymes (CAZymes), and antibiotic resistance genes (ARGs) were analyzed using comprehensive bioinformatic pipelines. Our analysis showed that diet profoundly shaped the gut microbiome architecture, with natural plant substrates supporting significantly higher microbial diversity than artificial diet. Proteobacteria dominated across all groups, with Enterobacteriaceae as the predominant family and Enterobacter as the most abundant genus. Notably, Enterobacter cancerogenus was consistently present in all dietary groups. Linear discriminant analysis effect size (LEfSe) identified diet-specific distinct microbial biomarkers: N. tabacum-fed larvae showed enrichment of Enterococcus spp. (Lactobacillales: Enterococcaceae), potentially facilitating alkaloid detoxification, while A. sinensis-fed larvae exhibited remarkable fungal diversity, particularly Ascomycota spanning multiple classes. Functional analysis revealed diet-specific enzymatic specialization, with A. sinensis-fed larvae exhibiting diverse CAZyme profiles including polysaccharide lyases and lignin-modifying enzymes. A. argyi-fed larvae showed moderate CAZyme abundance, while simplified artificial diet reduced both microbial diversity and functional complexity. KEGG pathway analysis showed N. tabacum-fed larvae exhibited comparatively enhanced oxidative metabolism with enriched respiratory chain components and cellular transport proteins, indicating metabolic adaptation to alkaloid detoxification. Co-occurrence analysis identified significant correlations between ARGs and specific microbial genera (Enterobacter, Escherichia, Enterococcus), suggesting potential resistance gene reservoirs within the gut microbiome. Our findings demonstrate that dietary substrate drives distinct microbial adaptations underlying the polyphagous lifestyle of L. serricorne, with diet-specific dependencies offering targets for microbiome-based biological control strategies.
{"title":"Metagenomic insights into diet-mediated gut microbial plasticity and functional adaptations in the tobacco beetle Lasioderma serricorne (Coleoptera: Ptinidae)","authors":"Junaid Zafar , Shuhao Zhang , Zhanpeng Zhu , Ziling Wan , Yanqing Li , Ang Li , Ling Huang , Xuewei Zhang , Chen Peng , Shuai Zhong , Fengliang Jin , Xiaoxia Xu","doi":"10.1016/j.jspr.2025.102931","DOIUrl":"10.1016/j.jspr.2025.102931","url":null,"abstract":"<div><div>The tobacco beetle, <em>Lasioderma serricorne</em>, is a globally distributed pest that inflicts major economic losses on stored products due its exceptional dietary adaptability. Understanding host-microbe-diet interactions is crucial for developing sustainable biological control strategies targeting insect-associated microorganisms. Here, we employed comparative metagenomic sequencing to characterize gut microbial communities in <em>L. serricorne</em> larvae reared on four dietary treatments i.e., artificial diet; <em>Artemisia argyi</em>; <em>Nicotiana tabacum</em> and <em>Angelica sinensis</em>. Taxonomic diversity, functional profiles, carbohydrate-active enzymes (CAZymes), and antibiotic resistance genes (ARGs) were analyzed using comprehensive bioinformatic pipelines. Our analysis showed that diet profoundly shaped the gut microbiome architecture, with natural plant substrates supporting significantly higher microbial diversity than artificial diet. Proteobacteria dominated across all groups, with Enterobacteriaceae as the predominant family and <em>Enterobacter</em> as the most abundant genus. Notably, <em>Enterobacter cancerogenus</em> was consistently present in all dietary groups. Linear discriminant analysis effect size (LEfSe) identified diet-specific distinct microbial biomarkers: <em>N. tabacum</em>-fed larvae showed enrichment of <em>Enterococcus</em> spp. (Lactobacillales: Enterococcaceae), potentially facilitating alkaloid detoxification, while <em>A. sinensis</em>-fed larvae exhibited remarkable fungal diversity, particularly Ascomycota spanning multiple classes. Functional analysis revealed diet-specific enzymatic specialization, with <em>A. sinensis</em>-fed larvae exhibiting diverse CAZyme profiles including polysaccharide lyases and lignin-modifying enzymes. <em>A. argyi</em>-fed larvae showed moderate CAZyme abundance, while simplified artificial diet reduced both microbial diversity and functional complexity. KEGG pathway analysis showed <em>N. tabacum</em>-fed larvae exhibited comparatively enhanced oxidative metabolism with enriched respiratory chain components and cellular transport proteins, indicating metabolic adaptation to alkaloid detoxification. Co-occurrence analysis identified significant correlations between ARGs and specific microbial genera (<em>Enterobacter</em>, <em>Escherichia</em>, <em>Enterococcus</em>), suggesting potential resistance gene reservoirs within the gut microbiome. Our findings demonstrate that dietary substrate drives distinct microbial adaptations underlying the polyphagous lifestyle of <em>L. serricorne</em>, with diet-specific dependencies offering targets for microbiome-based biological control strategies.</div></div>","PeriodicalId":17019,"journal":{"name":"Journal of Stored Products Research","volume":"116 ","pages":"Article 102931"},"PeriodicalIF":2.7,"publicationDate":"2025-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145836706","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-22DOI: 10.1016/j.jspr.2025.102940
Rosana Santos de Moraes , Nairiane dos Santos Bilhalva , Marisa Menezes Leal , Dthenifer Cordeiro Santana , Paulo Eduardo Teodoro , Larissa Pereira Ribeiro Teodoro , Paulo Carteri Coradi
the use of post-harvest technologies ensures quality, food safety, and efficiency in the production chain. Thus, this study aimed to evaluate a new method for classifying the physicochemical quality of flint and soft corn kernels using VNIR-SWIR spectroscopy and machine learning models. The results demonstrated that both the physical classification of kernel quality and the group of corn kernels significantly influenced the physicochemical composition. These results reinforce that, in addition to commercial classification, the type of endosperm influenced the physicochemical composition of the kernels, which may impact their industrial destination and added value. The hyperspectral curve showed that the lower the defect content in the sample, the lower the reflectance, a pattern observed for both soft and flint corn kernels. Despite the spectral differences between the groups, there are partially similar variation patterns, which reinforces the potential of using the hyperspectral curve combined with PCA for discrimination and evaluation of kernel quality. The confusion matrices confirmed the high hit rate in all classes. Thus, NIRS, combined with hyperspectral sensor analysis, was effective in assessing kernel quality, allowing for the identification of patterns related to defects. The support vector machine (SVM) obtained the best performance for the classification of soft and flint corn kernels, due to its high precision in accuracy metrics with values very close to 100 % for CC, 0.98 for Kappa, and 0.99 for F-score, being the most suitable alternative for the classifications. Thus, the adoption of non-destructive technology with indirect analysis in kernel storage units offers advantages over the traditional kernel classification method, including reduced analysis time, increased accuracy, the elimination of subjectivity, and a higher value of the final product. Its implementation contributes to segregation lots in the storage silos and optimize the operational efficiency, increasing traceability, and compliance with the quality standards required by the market.
{"title":"Classification process of the physicochemical quality of flint and soft corn kernels using VNIR-SWIR spectroscopy and machine learning: a technique for decision-making on batch segregation in corn storage and processing units","authors":"Rosana Santos de Moraes , Nairiane dos Santos Bilhalva , Marisa Menezes Leal , Dthenifer Cordeiro Santana , Paulo Eduardo Teodoro , Larissa Pereira Ribeiro Teodoro , Paulo Carteri Coradi","doi":"10.1016/j.jspr.2025.102940","DOIUrl":"10.1016/j.jspr.2025.102940","url":null,"abstract":"<div><div>the use of post-harvest technologies ensures quality, food safety, and efficiency in the production chain. Thus, this study aimed to evaluate a new method for classifying the physicochemical quality of flint and soft corn kernels using VNIR-SWIR spectroscopy and machine learning models. The results demonstrated that both the physical classification of kernel quality and the group of corn kernels significantly influenced the physicochemical composition. These results reinforce that, in addition to commercial classification, the type of endosperm influenced the physicochemical composition of the kernels, which may impact their industrial destination and added value. The hyperspectral curve showed that the lower the defect content in the sample, the lower the reflectance, a pattern observed for both soft and flint corn kernels. Despite the spectral differences between the groups, there are partially similar variation patterns, which reinforces the potential of using the hyperspectral curve combined with PCA for discrimination and evaluation of kernel quality. The confusion matrices confirmed the high hit rate in all classes. Thus, NIRS, combined with hyperspectral sensor analysis, was effective in assessing kernel quality, allowing for the identification of patterns related to defects. The support vector machine (SVM) obtained the best performance for the classification of soft and flint corn kernels, due to its high precision in accuracy metrics with values very close to 100 % for CC, 0.98 for Kappa, and 0.99 for F-score, being the most suitable alternative for the classifications. Thus, the adoption of non-destructive technology with indirect analysis in kernel storage units offers advantages over the traditional kernel classification method, including reduced analysis time, increased accuracy, the elimination of subjectivity, and a higher value of the final product. Its implementation contributes to segregation lots in the storage silos and optimize the operational efficiency, increasing traceability, and compliance with the quality standards required by the market.</div></div>","PeriodicalId":17019,"journal":{"name":"Journal of Stored Products Research","volume":"116 ","pages":"Article 102940"},"PeriodicalIF":2.7,"publicationDate":"2025-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145836707","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Understanding the olfactory recognition mechanisms underlying peanut-derived volatile organic compounds (VOCs) that mediate oviposition preference in Plodia interpunctella is essential for developing attractant-based pest control strategies. We identified a general odorant-binding protein gene, PintGOBP2, highly expressed in female antennae. It was expressed in a prokaryotic system, and fluorescence competitive binding assays with 15 peanut-derived VOCs revealed the strong binding affinities to octanal, 1-octen-3-ol, and acetophenone. Y-tube olfactometer assays confirmed that females were significantly attracted to these compounds. RNA interference (RNAi) silencing of PintGOBP2 reduced its expression level by 64.44 % after 24 h, resulting in markedly decreased EAGresponse, attraction and oviposition preference for the three VOCs. Molecular docking analysis showed that 1-octen-3-ol interacted mainly through hydrophobic contacts, while octanal and acetophenone formed both hydrophobic and hydrogen bonding interactions with PintGOBP2. These findings provide a foundation for semiochemical-based strategies in stored-product pest management.
{"title":"Olfactory recognition of peanut-derived volatiles mediated by odorant-binding protein PintGOBP2 in Plodia interpunctella","authors":"Chen Wang, Dianxuan Wang, Fangfang Zeng, Liang Chen","doi":"10.1016/j.jspr.2025.102928","DOIUrl":"10.1016/j.jspr.2025.102928","url":null,"abstract":"<div><div>Understanding the olfactory recognition mechanisms underlying peanut-derived volatile organic compounds (VOCs) that mediate oviposition preference in <em>Plodia interpunctella</em> is essential for developing attractant-based pest control strategies. We identified a general odorant-binding protein gene, <em>PintGOBP2</em>, highly expressed in female antennae. It was expressed in a prokaryotic system, and fluorescence competitive binding assays with 15 peanut-derived VOCs revealed the strong binding affinities to octanal, 1-octen-3-ol, and acetophenone. Y-tube olfactometer assays confirmed that females were significantly attracted to these compounds. RNA interference (RNAi) silencing of <em>PintGOBP2</em> reduced its expression level by 64.44 % after 24 h, resulting in markedly decreased EAGresponse, attraction and oviposition preference for the three VOCs. Molecular docking analysis showed that 1-octen-3-ol interacted mainly through hydrophobic contacts, while octanal and acetophenone formed both hydrophobic and hydrogen bonding interactions with PintGOBP2. These findings provide a foundation for semiochemical-based strategies in stored-product pest management.</div></div>","PeriodicalId":17019,"journal":{"name":"Journal of Stored Products Research","volume":"116 ","pages":"Article 102928"},"PeriodicalIF":2.7,"publicationDate":"2025-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145797505","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-20DOI: 10.1016/j.jspr.2025.102936
Asrat Otoru Oshone, Admasu Fanta Worku
Niger seed cake, a byproduct of oil extraction from niger seeds, is a protein and energy-rich livestock feed. However, it is prone to mold and mycotoxin contamination, which can harm animal health and the human food chain. This study evaluated the effectiveness of three storage bags in preventing mold growth and development of total aflatoxins (AFT) and ochratoxin A (OTA) in niger seed cake during three months of storage. Three storage bags were used in the experiment: Purdue Improved Crop Storage (PICS), Super GrainPro (SGP), and polypropylene (PPB), with each replicated three times using a completely randomized design. Mold, AFT, OTA, moisture, and nutrient contents were measured at baseline and after three months. During three months of storage, the average ambient temperature and, relative humidity were 23.8 ± 2.4 °C and 76.5 ± 3.1 % respectively. Enumeration of molds was performed using the spread plate method on potato dextrose agar. Mycotoxins were analyzed using the lateral flow immunoassay techniques on an indirect competitive immunoassay format. Data were analyzed at the 5 % significance level using one-way analysis of variance to evaluate storage bags over three months, and t-tests were used to compare baseline values with those at three months for mold, AFT, OTA, and nutritional composition. Mold and mycotoxins formation correlated with the nutritional composition of niger seed cake. Niger seed cake stored in PPB showed higher mold and mycotoxin levels than the baseline value. During three months of storage in PPB, AFT and OTA increased in niger seed cake by 4.42- and 1.70-fold, respectively. Except for ash, the nutritional composition of niger seed cake stored in PPB for 3 months decreased significantly from baseline. PICS and SGP bags significantly prevented mold and mycotoxin contamination, ensuring safer niger seed cake storage for farmers, oil millers, and livestock industries.
{"title":"Comparative evaluation of storage bags for mold and mycotoxins control in niger seed cake","authors":"Asrat Otoru Oshone, Admasu Fanta Worku","doi":"10.1016/j.jspr.2025.102936","DOIUrl":"10.1016/j.jspr.2025.102936","url":null,"abstract":"<div><div>Niger seed cake, a byproduct of oil extraction from niger seeds, is a protein and energy-rich livestock feed. However, it is prone to mold and mycotoxin contamination, which can harm animal health and the human food chain. This study evaluated the effectiveness of three storage bags in preventing mold growth and development of total aflatoxins (AFT) and ochratoxin A (OTA) in niger seed cake during three months of storage. Three storage bags were used in the experiment: Purdue Improved Crop Storage (PICS), Super GrainPro (SGP), and polypropylene (PPB), with each replicated three times using a completely randomized design. Mold, AFT, OTA, moisture, and nutrient contents were measured at baseline and after three months. During three months of storage, the average ambient temperature and, relative humidity were 23.8 ± 2.4 °C and 76.5 ± 3.1 % respectively. Enumeration of molds was performed using the spread plate method on potato dextrose agar. Mycotoxins were analyzed using the lateral flow immunoassay techniques on an indirect competitive immunoassay format. Data were analyzed at the 5 % significance level using one-way analysis of variance to evaluate storage bags over three months, and t-tests were used to compare baseline values with those at three months for mold, AFT, OTA, and nutritional composition. Mold and mycotoxins formation correlated with the nutritional composition of niger seed cake. Niger seed cake stored in PPB showed higher mold and mycotoxin levels than the baseline value. During three months of storage in PPB, AFT and OTA increased in niger seed cake by 4.42- and 1.70-fold, respectively. Except for ash, the nutritional composition of niger seed cake stored in PPB for 3 months decreased significantly from baseline. PICS and SGP bags significantly prevented mold and mycotoxin contamination, ensuring safer niger seed cake storage for farmers, oil millers, and livestock industries.</div></div>","PeriodicalId":17019,"journal":{"name":"Journal of Stored Products Research","volume":"116 ","pages":"Article 102936"},"PeriodicalIF":2.7,"publicationDate":"2025-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145836678","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-20DOI: 10.1016/j.jspr.2025.102935
Sreelakshmi Mullachery, Naduvilthara U. Visakh, Berin Pathrose
Psidium guajava (Myrtaceae), commonly referred to as guava, is a tropical plant widely cultivated across the tropics and subtropics for its fruits. The plant is also characterized by anti-inflammatory, anti-microbial, anti-oxidant and anti-diabetic properties. However, the training and pruning waste of these trees is often discarded as agricultural waste. This study was conducted to extract essential oil from dried, ground guava leaves through hydro-distillation, identify its chemical constituents through GC-MS analysis, and investigate the potential of using guava leaf oil as a biopesticide against three key stored grain pests, Tribolium castaneum, Sitophilus oryzae and Rhyzopertha dominica. The leaves yielded 1.4 ± 0.01 % oil, caryophyllene (24.43 %), D-limonene (16.13 %) and α-bisabolene (8.13 %) were identified as the major chemical components and the oil demonstrated excellent contact (LC50 at 24 h was 7.92 mg/cm2 for T. castaneum, 12.71 mg/cm2 for S. oryzae and 5.41 mg/cm2 for R. dominica), fumigant (LC50 at 24 h was 18.91 μL/L air for T. castaneum, 33.25 μL/L air for S. oryzae and 11.25 μL/L air for R. dominica) and repellent activity. The highest mean repellency was exhibited against T. castaneum at 0.09 % across varying exposure periods. An area preference technique was used to evaluate the repellent activity. Additionally, a phytotoxicity test on wheat seeds showed no significant effect on seed germination or seedling growth. Thus, guava leaf waste generated during training and pruning can be used to extract leaf oil, which can then be used as a safe botanical pesticide to control storage pests.
番石榴（番石榴科），通常被称为番石榴，是一种热带植物，因其果实而广泛种植于热带和亚热带地区。该植物还具有抗炎、抗微生物、抗氧化和抗糖尿病的特性。然而，这些树木的训练和修剪废料往往被作为农业废物丢弃。本研究采用水蒸气蒸馏法提取干燥磨碎的番石榴叶精油，并通过气相色谱-质谱分析鉴定其化学成分，探讨番石榴叶精油作为生物农药对三种主要储粮害虫——castaneum、Sitophilus oryzae和Rhyzopertha dominica的应用潜力。叶片出油率为1.4±0.01%，主要化学成分为石蜡烯（24.43%）、d -柠檬烯（16.13%）和α-双abolene(8.13%)，精油具有良好的接触性（24 h LC50为7.92 mg/cm2、12.71 mg/cm2和5.41 mg/cm2）、熏蒸剂（24 h LC50为18.91 μL/L空气、33.25 μL/L空气和11.25 μL/L空气）和驱避活性。在不同的暴露时间内，平均驱避率最高，为0.09%。采用区域偏好法评价其驱避活性。此外，小麦种子的植物毒性试验表明，对种子萌发和幼苗生长没有显著影响。因此，在训练和修剪过程中产生的番石榴叶废料可以用来提取叶油，然后可以用作安全的植物性农药来控制储存害虫。
{"title":"Valorization of guava leaf waste: Chemical characterization and comparative bioefficacy of Psidium guajava L. essential oil against key stored-product insect pests","authors":"Sreelakshmi Mullachery, Naduvilthara U. Visakh, Berin Pathrose","doi":"10.1016/j.jspr.2025.102935","DOIUrl":"10.1016/j.jspr.2025.102935","url":null,"abstract":"<div><div><em>Psidium guajava</em> (Myrtaceae), commonly referred to as guava, is a tropical plant widely cultivated across the tropics and subtropics for its fruits. The plant is also characterized by anti-inflammatory, anti-microbial, anti-oxidant and anti-diabetic properties. However, the training and pruning waste of these trees is often discarded as agricultural waste. This study was conducted to extract essential oil from dried, ground guava leaves through hydro-distillation, identify its chemical constituents through GC-MS analysis, and investigate the potential of using guava leaf oil as a biopesticide against three key stored grain pests<em>, Tribolium castaneum, Sitophilus oryzae</em> and <em>Rhyzopertha dominica.</em> The leaves yielded 1.4 ± 0.01 % oil, caryophyllene (24.43 %), D-limonene (16.13 %) and α-bisabolene (8.13 %) were identified as the major chemical components and the oil demonstrated excellent contact (LC<sub>50</sub> at 24 h was 7.92 mg/cm<sup>2</sup> for <em>T. castaneum,</em> 12.71 mg/cm<sup>2</sup> for <em>S. oryzae</em> and 5.41 mg/cm<sup>2</sup> for <em>R. dominica</em>), fumigant (LC<sub>50</sub> at 24 h was 18.91 μL/L air for <em>T. castaneum,</em> 33.25 μL/L air for <em>S. oryzae</em> and 11.25 μL/L air for <em>R. dominica</em>) and repellent activity. The highest mean repellency was exhibited against <em>T. castaneum</em> at 0.09 % across varying exposure periods. An area preference technique was used to evaluate the repellent activity. Additionally, a phytotoxicity test on wheat seeds showed no significant effect on seed germination or seedling growth. Thus, guava leaf waste generated during training and pruning can be used to extract leaf oil, which can then be used as a safe botanical pesticide to control storage pests.</div></div>","PeriodicalId":17019,"journal":{"name":"Journal of Stored Products Research","volume":"116 ","pages":"Article 102935"},"PeriodicalIF":2.7,"publicationDate":"2025-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145797503","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-18DOI: 10.1016/j.jspr.2025.102932
Li Hou, Mingliang Liu, Yanfen Liu, Xiaodong Tan, Chunxiang Cheng, Wentao Sun, Nan Tang
Accurate detection of pests of different sizes is essential for smart monitoring of stored grain. However, mainstream deep learning models like YOLO11 and YOLOv8 often fail to detect tiny pests in complex storage settings—largely due to insufficient retention of low-level features—and their high computational demands restrict deployment on edge devices with limited resources. This study addresses these challenges by proposing an improved model, ABF-YOLO11, targeting three common stored-grain pests: Sitophilus granarius, Rhyzopertha dominica, and Tribolium castaneum. The model's innovation lies in integrating three key components to balance performance and lightweight design: the ADown module (preserves fine-grained features during downsampling while cutting parameters), BiFPN (boosts multi-scale pest detection via weighted bidirectional feature fusion), and the Focaler-MPDIoU loss function (enhances recognition of hard-to-classify samples and bounding box accuracy). On a stored-grain pest dataset, ABF-YOLO11 achieves 0.944 mAP50 and 0.923 precision—2.0 % and 4.1 % higher than YOLO11—with only 1.50 M parameters and 5.0 GFLOPs (42.0 % fewer parameters and 20.6 % lower computational cost than YOLO11). It maintains stable performance in complex backgrounds, excels at detecting tiny targets and distinguishing similar pest species, and can be seamlessly integrated into smart granary systems (e.g., edge sensors, mobile inspection devices) for real-time, low-power monitoring, helping reduce grain loss caused by pests.
{"title":"ABF-YOLO11: A multi-scale adaptive detection and lightweight recognition model for stored-grain pests","authors":"Li Hou, Mingliang Liu, Yanfen Liu, Xiaodong Tan, Chunxiang Cheng, Wentao Sun, Nan Tang","doi":"10.1016/j.jspr.2025.102932","DOIUrl":"10.1016/j.jspr.2025.102932","url":null,"abstract":"<div><div>Accurate detection of pests of different sizes is essential for smart monitoring of stored grain. However, mainstream deep learning models like YOLO11 and YOLOv8 often fail to detect tiny pests in complex storage settings—largely due to insufficient retention of low-level features—and their high computational demands restrict deployment on edge devices with limited resources. This study addresses these challenges by proposing an improved model, ABF-YOLO11, targeting three common stored-grain pests: <em>Sitophilus granarius</em>, <em>Rhyzopertha dominica</em>, and <em>Tribolium castaneum</em>. The model's innovation lies in integrating three key components to balance performance and lightweight design: the ADown module (preserves fine-grained features during downsampling while cutting parameters), BiFPN (boosts multi-scale pest detection via weighted bidirectional feature fusion), and the Focaler-MPDIoU loss function (enhances recognition of hard-to-classify samples and bounding box accuracy). On a stored-grain pest dataset, ABF-YOLO11 achieves 0.944 mAP50 and 0.923 precision—2.0 % and 4.1 % higher than YOLO11—with only 1.50 M parameters and 5.0 GFLOPs (42.0 % fewer parameters and 20.6 % lower computational cost than YOLO11). It maintains stable performance in complex backgrounds, excels at detecting tiny targets and distinguishing similar pest species, and can be seamlessly integrated into smart granary systems (e.g., edge sensors, mobile inspection devices) for real-time, low-power monitoring, helping reduce grain loss caused by pests.</div></div>","PeriodicalId":17019,"journal":{"name":"Journal of Stored Products Research","volume":"116 ","pages":"Article 102932"},"PeriodicalIF":2.7,"publicationDate":"2025-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145797504","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号