Agricultural Production Environment Research Articles

Study on the Influence of Agricultural Scale Management Mode on Production Efficiency Based on Meta-Analysis

Agricultural scale management is an important means of improving agricultural production efficiency. To answer the controversy over whether different agricultural scale management models can improve production efficiency, this paper obtained 117,627 samples from 68 domestic and foreign literature and used meta-analysis to study the effects of land-scale operation (LSO) and service-scale operation (SSO) on agricultural production efficiency (APE). The moderators that affect the relationship between LSO, SSO, and APE are also examined. The research results show that (1) both LSO and SSO positively impact production efficiency, but LSO has a stronger impact. (2) The relationship between LSO and APE is positively regulated by the agricultural production and operation environment, characteristics of the agricultural location, degree of farmers’ participation, and types of APE, and negatively regulated by the research situation. (3) The relationship between SSO and APE is negatively regulated by APE types and data types. Therefore, the government should promote agricultural dual-scale operation according to the local conditions through the complementary advantages of LSO and SSO, so as to fully release the promotion potential of APE.

Monitoring of Extreme Drought in the Yangtze River Basin in 2022 Based on Multi-Source Remote Sensing Data

The Yangtze River Basin experienced a once-in-a-century extreme drought in 2022 due to extreme weather, which had a serious impact on the local agricultural production and ecological environment. In order to investigate the spatial distribution and occurrence of the extreme drought events, this study used multi-source remote sensing data to monitor the extreme drought events in the Yangtze River Basin in 2022. In this study, the gravity satellite data product CSR_Mascon was used to calculate the GRACE Drought Intensity Index (GRACE-DSI), which was analyzed and compared with the commonly used meteorological drought indices, relative soil humidity, and soil water content data. The results show that (1) terrestrial water storage change data can well reflect the change in water storage in the Yangtze River Basin. Throughout the year, the average change in terrestrial water storage in the Yangtze River Basin from January to June is higher than the average value of 33.47 mm, and the average from July to December is lower than the average value of 48.17 mm; (2) the GRACE-DSI responded well to the intensity and spatial distribution of drought events in the Yangtze River Basin region in 2022. From the point of view of drought area, the Yangtze River Basin showed a trend of extreme drought increasing first, and then decreasing in the area of different levels of drought, and the range of drought reached a maximum in September with a drought area of 175.87 km2, which accounted for 97.71 per cent of the total area; at the same time, the area of extreme drought was the largest, with an area of 85.69 km2; (3) the spatial and temporal variations of the GRACE-DSI and commonly used meteorological drought indices were well correlated, with correlation coefficients above 0.750, among which the correlation coefficient of the SPEI-3 was higher at 0.937; (4) the soil moisture and soil relative humidity products from the CLDAS, combined with soil moisture products from the GLDAS, reflect the starting and ending times of extreme drought events in the Yangtze River Basin in 2022 well, using the information from the actual stations. In conclusion, gravity satellite data, analyzed in synergy with data from multiple sources, help decision makers to better understand and respond to drought.

Improved YOLOv8 and SAHI Model for the Collaborative Detection of Small Targets at the Micro Scale: A Case Study of Pest Detection in Tea

Pest target identification in agricultural production environments is challenging due to the dense distribution, small size, and high density of pests. Additionally, changeable environmental lighting and complex backgrounds further complicate the detection process. This study focuses on enhancing the recognition performance of tea pests by introducing a lightweight pest image recognition model based on the improved YOLOv8 architecture. First, slicing-aided fine-tuning and slicing-aided hyper inference (SAHI) are proposed to partition input images for enhanced model performance on low-resolution images and small-target detection. Then, based on an ELAN, a generalized efficient layer aggregation network (GELAN) is designed to replace the C2f module in the backbone network, enhance its feature extraction ability, and construct a lightweight model. Additionally, the MS structure is integrated into the neck network of YOLOv8 for feature fusion, enhancing the extraction of fine-grained and coarse-grained semantic information. Furthermore, the BiFormer attention mechanism, based on the Transformer architecture, is introduced to amplify target characteristics of tea pests. Finally, the inner-MPDIoU, based on auxiliary borders, is utilized as a replacement for the original loss function to enhance its learning capacity for complex pest samples. Our experimental results demonstrate that the enhanced YOLOv8 model achieves a precision of 96.32% and a recall of 97.95%, surpassing those of the original YOLOv8 model. Moreover, it attains an mAP@50 score of 98.17%. Compared to Faster R-CNN, SSD, YOLOv5, YOLOv7, and YOLOv8, its average accuracy is 17.04, 11.23, 5.78, 3.75, and 2.71 percentage points higher, respectively. The overall performance of YOLOv8 outperforms that of current mainstream detection models, with a detection speed of 95 FPS. This model effectively balances lightweight design with high accuracy and speed in detecting small targets such as tea pests. It can serve as a valuable reference for the identification and classification of various insect pests in tea gardens within complex production environments, effectively addressing practical application needs and offering guidance for the future monitoring and scientific control of tea insect pests.

Smart Irrigation and Pesticides System for Enhancing Harvesting in Agriculture Using IoT

Agriculture plays a vital role in the economy of many countries, serving as a primary source of food. However, traditional farming practices often face challenges such as inefficient water usage, over-reliance on pesticides, and inconsistent monitoring of crop conditions. In recent years, the Internet of Things (IoT) has emerged as a promising technology to address these challenges by providing real-time monitoring and control of agricultural processes. This leads in excessive water use, which has an impact on agricultural output. This integrated system means that farmers only need to perform little manual monitoring. The opening and closing of the reservoir and field gates, as well as the verification of the plant's age and the matching optimal water level necessary for that time of plant age, are all automated. Farmers may use their mobile phones to check the water level in the field since they are continually receiving updates on the field work. The most labour-intensive and time consuming activity in agriculture is water management. In our project, the water level in the agriculture harvest sting is measured, and if it is lower than the required level, the paddy field is automatically irrigated for the water source or tank using a water pump. It can comprehensively monitor air temperature and humidity, light intensity, ultraviolet light and other environmental elements. In specific use, the agricultural production environment monitoring system starts from the actual needs of the plantation zone and is assembled according to the environmental elements it needs to monitor.

Analysis of Image Data Augmentation in Smart Agriculture

The demand for agricultural production automation in smart agriculture has made deep learning widely used in tasks, e.g., disease identification, weed-crop segment and fruit detection based on image processing. However, in agricultural production environments, there are limitations in data acquisition, so data enhancement technology has also been widely used. This study introduces two data enhancement methods commonly used in smart agriculture, namely image transformation and GAN-based, and analyzes their specific applications and roles in three common task scenarios: disease identification, weed-crop segment, and fruit detection. According to the analysis, the data enhancement is mainly used to deal with class imbalance problems and over-fitting problems, and can significantly improve model performance in specific scenarios; the complexity of tasks and image structures has an impact on the application effects of different methods, and image segmentation tasks. It is necessary to increase the number of image modes, so methods such as GAN-based are adopted. However, due to the complexity of the image, fruit detection uses simple transformation methods to achieve better results. In addition, data enhancement methods designed for specific tasks, such as simulating synthetic images, can also achieve good results. However, data augmentation technology still has mode collapse problems and is too cumbersome in its application, which points out the direction for subsequent research. Overall, this research sorts out the applications of common data enhancement methods, analyzes the characteristics and limitations of these methods, and hopes to provide guidance for subsequent research.

Organic Remobilization of zinc and phosphorus availability to plants by application of mineral solubilizing bacteria Pseudomonas aeruginosa

Incessant utilization of chemical fertilizers leads to the accumulation of minerals in the soil, rendering them unavailable to plants. Unaware of the mineral reserves present in the soil, farming communities employ chemical fertilizers once during each cultivation, a practice that causes elevated levels of insoluble minerals within the soil. The use of biofertilizers on the other hand, reduces the impact of chemical fertilizers through the action of microorganisms in the product, which dissolves minerals and makes them readily available for plant uptake, helping to create a sustainable environment for continuous agricultural production. In the current investigation, a field trial employing Arachis hypogaea L was conducted to evaluate the ability of Pseudomonas aeruginosa to enhance plant growth and development by solubilizing minerals present in the soil (such as zinc and phosphorus). A Randomized Complete Block Design (RCBD) included five different treatments as T1: Un inoculated Control; T2: Seeds treated with a liquid formulation of P. aeruginosa; T3: Seeds treated with a liquid formulation of P. aeruginosa and the soil amended with organic manure (farmyard); T4: Soil amended with organic manure (farmyard) alone; T5: Seeds treated with lignite (solid) based formulation of P. aeruginosa were used for the study. Efficacy was determined based on the plant's morphological characters and mineral contents (Zn and P) of plants and soil. Survival of P. aeruginosa in the field was validated using Antibiotic Intrinsic patterns (AIP). The results indicated that the combination treatment of P. aeruginosa liquid formulation and organic fertilizer (farmyard) (T3) produced the highest biometric parameters and mineral (Zn and P) content of the groundnut plants and the soil. This outcome is likely attributed to the mineral solubilizing capability of P. aeruginosa. Furthermore, the presence of farmyard manure increased the metabolic activity of P. aeruginosa by inducing its heterotrophic activity, leading to higher mineral content in T3 soil compared to other soil treatments. The AIP data confirmed the presence of the applied liquid inoculant by exhibiting a similar intrinsic pattern between the in vitro isolate and the isolate obtained from the fields. In summary, the Zn and P solubilization ability of P. aeruginosa facilitates the conversion of soil-unavailable mineral form into a form accessible to plants. It further proposes the utilization of the liquid formulation of P. aeruginosa as a viable solution to mitigate the challenges linked to solid-based biofertilizers and the reliance on mineral-based chemical fertilizers.

Discriminative semisupervised dictionary learning method with graph embedding and pairwise constraints for crop disease image recognition

Crop diseases are serious biological disasters that affect agricultural production and ecological environment safety. Accurate detection and identification of disease types can effectively reduce the harm caused by diseases. However, there are various types of crop diseases and labeling them consumes a lot of labor and material resources. A discriminative semisupervised dictionary learning method with graph embedding and pairwise constraints (DSDL-GEPC) is proposed in this study. DSDL-GEPC adopts semisupervised dictionary learning as the baseline model, which combines the reconstruction error of all training samples and the classification error of labeled training samples. DSDL-GEPC consists of three learning terms. The graph embedding term is constructed based on dictionary atoms to obtain global geometric structural information of data samples. Using the pairwise constraint information of similar/heterogeneous images, the pairwise constraints term is constructed to enhance the dictionary's discriminability in terms of within-class similarity and between-class separability. In addition, the relaxed low-rank classification term is constructed, which introduces the ε-draggings technique and low-rank constraint on the relaxed label matrix to learn a discriminative classifier. Experiments are performed to compare the classification performance of DSDL-GEPC on the PlantVillage dataset. The results validate the effectiveness of DSDL-GEPC for crop disease image recognition.

Effects of Reclaimed Water Irrigation on Grain Quality and Endogenous Estrogen Concentrations of Winter Wheat

Reclaimed water irrigation can effectively alleviate the shortage of water resources in arid and semi-arid areas; however, reclaimed water contains organic pollutants that may enter the agricultural production environment through irrigation, such as endogenous estrogens, so people have always paid attention to the safety of reclaimed water irrigation. In this paper, we studied the effects of reclaimed water irrigation, groundwater irrigation, and alternating irrigation of the two water sources on grain quality, as well as endogenous estrogen concentrations of winter wheat and topsoil in the North China Plain during 2015–2016. The results show that the concentrations of crude protein, total soluble sugar, crude ash, crude starch, and reduction-type vitamin C (reduction-type VC) in the winter wheat grains were 12.5–16.4%, 0.85–2.15%, 1.85–3.28%, 61.5–75.0%, and 4.9–16.0 mg/kg, respectively. There were no significant differences in these quality indexes among the irrigation treatments (p > 0.05). The concentrations of endogenous estrogen in the surface soil and winter wheat grain under all irrigation treatments were 0.34–4.01 μg/kg and below the limits of detection (ND)–3.71 μg/kg, respectively. There were no significant differences in the concentrations of endogenous estrogen in the soil and wheat grain among the different irrigation treatments (p > 0.05). The bioconcentration factor (BCF) of the endogenous estrogen in the soil–winter wheat system was 0.08–1.90, and there was no significant difference in the BCF among the irrigation treatments (p > 0.05). Compared with groundwater irrigation, reclaimed water irrigation did not significantly affect endogenous estrogen concentrations in the soil and winter wheat, as well as the bioconcentration factors.

Whole-genome sequencing of diverse wheat accessions uncovers genetic changes during modern breeding in China and the United States.

Breeding has dramatically changed the plant architecture of wheat (Triticum aestivum), resulting in the development of high-yielding varieties adapted to modern farming systems. However, how wheat breeding shaped the genomic architecture of this crop remains poorly understood. Here, we performed a comprehensive comparative analysis of a whole-genome resequencing panel of 355 common wheat accessions (representing diverse landraces and modern cultivars from China and the United States) at the phenotypic and genomic levels. The genetic diversity of modern wheat cultivars was clearly reduced compared to landraces. Consistent with these genetic changes, most phenotypes of cultivars from China and the United States were significantly altered. Of the 21 agronomic traits investigated, 8 showed convergent changes between the 2 countries. Moreover, of the 207 loci associated with these 21 traits, more than half overlapped with genomic regions that showed evidence of selection. The distribution of selected loci between the Chinese and American cultivars suggests that breeding for increased productivity in these 2 regions was accomplished by pyramiding both shared and region-specific variants. This work provides a framework to understand the genetic architecture of the adaptation of wheat to diverse agricultural production environments, as well as guidelines for optimizing breeding strategies to design better wheat varieties.

Chlorpyrifos induces placental oxidative stress and barrier dysfunction by inducing mitochondrial apoptosis through the ERK/MAPK signaling pathway: In vitro and in vivo studies

Chlorpyrifos (CPF) is an organophosphorus pesticide that is widely used in agricultural production and residential environments worldwide. In this study, we determined the harmful effects and toxicological mechanism of CPF in porcine trophectoderm (pTr) cells and the placenta of female mice during pregnancy. The findings revealed that CPF significantly decreased cell viability and increased intracellular lactate dehydrogenase (LDH) release in pTr cells. Similarly, CPF induced reproductive toxicity in pregnant maternal mice, including decreased maternal, fetal, and placental weights. Moreover, following CPF treatment, pTr cells and the placenta of female mice showed significant apoptosis. JC-1 staining and flow cytometry analysis also revealed that the mitochondrial membrane potential (MMP) of pTr cells treated with CPF was significantly depolarized. Additionally, CPF can induce an increase in reactive oxygen species (ROS) and barrier dysfunction in pTr cells and the placenta of female mice. We further verified that CPF-induced mitochondrial apoptosis is mediated by the MAPK signaling pathway, as shown by using of small molecular inhibitors of related proteins. Also, CPF-induced oxidative stress, barrier dysfunction, and mitochondrial apoptosis in pTr cells were alleviated by U0126, an inhibitor of the ERK/MAPK signaling pathway. These findings suggested that exposure to CPF in early pregnancy might be a potential risk fator affecting placental formation and function in humans and animals.

Identification and interpretation of internal factors influencing the agricultural insurance market in Ukraine

The functioning of the agricultural insurance market in Ukraine is an objective reality and a necessity due to the urgent need to create a safe environment for agricultural production. The ability to consolidate financial resources in the field of insurance contributes to increased responsibility for probable risks distributed in space and time, so it is worth considering insurance relations in the agricultural sector. The purpose of the study is to theoretically and methodologically justify the processes of identifying internal factors influencing the further functioning of the agricultural insurance market in Ukraine. The methodological perspective of this study is based on the use of general scientific and specific methods of cognitive activity: abstract-logical, economic-statistical, structuralfunctional, deduction, induction, and formalisation. A panel database was formed for the period 2012-2021 broken down by the regions of Ukraine. The territorial features of changes in the ratio between insured and sown areas in the western, central, eastern, and southern regions of Ukraine are reflected. Based on the results of the conducted modelling (two models are presented simultaneously: a regular model and a one-year lag model), the influence on the insured area of such factors as sown area, yield, gross harvest, and internal price for agricultural products is established. A comparative characteristic of the statistical qualities of the presented models was made using indicators such as “Residual standard error”, “Multiple R-squared”, and “Adjusted R-squared”. The results of the regression analysis showed the statistical significance of certain factors in motivating agricultural commodity producers to participate in the insurance process. The practical value of the proposed measures lies in formulating practical recommendations for enhancing cooperation among stakeholders in the agricultural insurance market during wartime and post-war periods

Effect of saline–alkaline stresses on the interspecific competition between Aegilops tauschii and Triticum aestivum

Soil salinization has produced severe consequences on global agricultural production and ecological environment. Based on the features of saline soils in China, through mixed NaCl, NaHCO3, Na2SO4, and Na2CO3 at varying ratios to simulate the salinity–alkalinity stress and conducted a controlled pot experiment using De Wit replacement method. The effects of salinity–alkalinity stress on the growth of Aegilops tauschii and its competition with wheat were explored to provide a reference for the study of invasion mechanism of A. tauschii. The result showed that, the salinity–alkalinity stress inhibited the growth and development of both the species, which was reflected in plant height, leaf area and total biomass indicators. Secondly, the tolerance of both plant species to salinity–alkalinity stress was improved by increasing the superoxide dismutase (SOD) activity and the proline content. However, as the salinity–alkalinity stress was exacerbated, the relative conductivity and thiobarbituric acid (TBARS) content increased significantly in both the species. As suggested by the level of increase in SOD activity, proline content, relative conductivity, and TBARS content, A. tauschii was more tolerant to the salinity–alkalinity stress than wheat. Finally, it can be seen from the value of the competition balance index, A. tauschii was still more competitive than wheat even under salinity–alkalinity stress. In summary, A. tauschii was more tolerant of the salinity–alkalinity stress than wheat through the favorable adjustment of morphology, biomass allocation pattern, and physiological features, which is expected to increase its invasion damage to wheat.

Drought monitoring based on temperature vegetation dryness index and its relationship with anthropogenic pressure in a subtropical humid watershed in China

Drought has serious impacts on agricultural production and ecological environment in China. Study on the causes of drought and drought monitoring technology is crucial for drought mitigation. This study conducts drought monitoring based on temperature vegetation dryness index (TVDI) and investigates with the impact of anthropogenic pressure in the Poyang Lake basin (PYLB). In the study, the MODIS normalized vegetation index (NDVI) and land surface temperature (LST) products from 2000 to 2021 were selected to construct NDVI-LST feature space processed with elevation correction, from which the monthly TVDI data was calculated. The land use data and nighttime light data were used to construct the anthropogenic pressure index (API). The results indicate that the frequency of drought in the PYLB is higher in the south and lower in the north. According to drought influence scope (DIS), the occurrence frequency of basin-wide drought (DIS ≥ 50%) in the PYLB is the highest, followed by regional drought (33%≤DIS < 50%), partial local drought (25%≤DIS < 33%), and local drought (10%≤DIS < 25%). The basin as a whole showed an obvious wetting trend in spring and a drying trend in autumn. During the study period, API in the north of PYLB was obviously higher than that in the south, and showed a general increasing trend. The contribution of human activities to drought in the PYLB increased exponentially with the improvement of social development level. Meanwhile, the impact of human activities on drought shows spatial differentiation. Ecological protection and green development have been implemented in the northerly region of the PYLB, and anthropogenic pressure inhibited the degree of drought. Oppositely, extensive socioeconomic development in the central and southern parts of the PYLB has caused damage to the environment and indirectly aggravated regional drought. This study makes important contributions to exploring the response between remote sensing drought monitoring and anthropogenic pressure and promoting regional drought mitigation in humid areas.

Soil Salinity Estimation in Cotton Fields in Arid Regions Based on Multi-Granularity Spectral Segmentation (MGSS)

Soil salinization seriously threatens agricultural production and ecological environments in arid areas. The accurate and rapid monitoring of soil salinity and its spatial variability is of great significance for the amelioration of saline soils. In this study, 191 soil samples were collected from cotton fields in southern Xinjiang, China, to obtain spectral reflectance and electrical conductivity (EC) indoors. Then, multi-granularity spectral segmentation (MGSS) and seven conventional spectral preprocessing methods were employed to preprocess the spectral data, followed by the construction of partial least squares regression (PLSR) models for soil EC estimation. Finally, the performance of the models was compared. The results showed that compared with conventional spectral preprocessing methods, MGSS could greatly improve the correlation between spectrum and soil EC, extract the weak spectral information of soil EC, and expand the spectral utilization range. The model validation results showed that the PLSR model based on the second-order derivative (2nd-der-PLSR) had the highest estimation accuracy among the models constructed by conventional methods. However, the PLSR model based on MGSS (MGSS-PLSR) had the highest estimation accuracy among all models, with Rp2 (0.901) and RPD (3.080) being 0.151 and 1.302 higher than those of the 2nd-der-PLSR model, respectively, and nRMSEP (5.857%) being 4.29% lower than that of the 2nd-der-PLSR model. The reason for the high accuracy of the MGSS-PLSR model is as follows: In the continuous segmentation of the raw spectrum by MGSS, the bands with strong and weak correlations with respect to soil EC were concentrated during low granularity segmentation. With the increase in granularity level, the spectral features decreased and were distributed discretely. In addition, the locations of spectral features were also different at different granularity levels. Therefore, the spectral features of soil EC can be effectively extracted by the MGSS, which significantly improves the spectral estimation accuracy of soil salinity. This study provides a new technical means for soil salinity estimation in arid areas.

A Novel LoRa-Based Platform for Remote Monitoring of Large-Scale Rice Fields

The utilization of wireless sensor networks is widespread across numerous industries, including smart agriculture, where the need to remotely monitor agricultural production environments on a large scale is imperative for improving farm product quality and preserving the ecosystem. In this paper, a novel and efficient wireless sensor network framework is presented, which employs LoRa technology to facilitate remote environmental monitoring during agricultural cultivation. Our proposed network framework possesses the capability of long-range data transmission, secure data transfer from source nodes to gateway, high reliability, and easy deployment for large-scale farm monitoring applications. The effectiveness of our model was evaluated in An Giang province, Vietnam, with monitoring indicators encompassing temperature, air humidity, soil moisture, pH of water, CH4, and NH3. The findings reveal that the wireless sensor network model for remote monitoring of the rice cultivation environment fulfills the requirements of good and complete data transmission, high data network security, and a long signal transmission distance of up to 3.5 km at a reasonable cost.

An environment safety monitoring system for agricultural production based on artificial intelligence, cloud computing and big data networks

Monitoring the agricultural production environment is crucial for optimal crop growth and resource efficiency. Cloud Computing, Artificial Intelligence (AI), and Big Data have revolutionized traditional agriculture, promising improved output and product quality. The popularity of these technologies drives their application in safety monitoring. This system facilitates data collection and transmission among equipment, overcoming challenges of traditional systems like investment, costs, and maintenance. In this paper, cloud computing-based AI optimization technology and big data network were proposed to monitor the safety of the agricultural production environment, and the shortcomings of traditional distance vector hop (DV hop) positioning algorithms were analyzed in depth. RSSI (Received Signal Strength Indication) technology improved the traditional DV Hop location method. The paper analyses direct and indirect transmission for data transmission between WSN and cloud nodes and favors indirect transmission because it consumes less invalid energy. Finally, the article compares several evaluations of alternative algorithms for monitoring system performance, including data transmission reliability, data reception rate, and data delay. The experimental results in this paper showed that in the data reception rate test, the data reception rate of System 2 was 97% at the lowest and 99% at the highest, both exceeding 95%.

Review on the Application of Agricultural Waste Resources

Globally, the positive effects of aerobic fermentation industry on the resource utilization of agricultural waste and products on the agricultural production environment have been fully proved. Therefore, it is of great and far-reaching significance to improve the development level of agricultural waste aerobic fermentation industry. In this paper, the current situation of agricultural waste in China, the requirements of aerobic fermentation raw materials and the process were reviewed in order to provide scientific basis for the study of aerobic fermentation of agricultural waste.

IBSA_Net: A Network for Tomato Leaf Disease Identification Based on Transfer Learning with Small Samples

Tomatoes are a crop of significant economic importance, and disease during growth poses a substantial threat to yield and quality. In this paper, we propose IBSA_Net, a tomato leaf disease recognition network that employs transfer learning and small sample data, while introducing the Shuffle Attention mechanism to enhance feature representation. The model is optimized by employing the IBMax module to increase the receptive field and adding the HardSwish function to the ConvBN layer to improve stability and speed. To address the challenge of poor generalization of models trained on public datasets to real environment datasets, we developed an improved PlantDoc++ dataset and utilized transfer learning to pre-train the model on PDDA and PlantVillage datasets. The results indicate that after pre-training on the PDDA dataset, IBSA_Net achieved a test accuracy of 0.946 on a real environment dataset, with an average precision, recall, and F1-score of 0.942, 0.944, and 0.943, respectively. Additionally, the effectiveness of IBSA_Net in other crops is verified. This study provides a dependable and effective method for recognizing tomato leaf diseases in real agricultural production environments, with the potential for application in other crops.

The abundance, characteristics and distribution of microplastics (MPs) in farmland soil—Based on research in China

Microplastics (MPs) in farmland soil deteriorate soil environment and increase food toxicity, thereby threatening the agricultural production environment and human safety. However, a systematic understanding of MPs pollution in farmland soil is lacking in China. Therefore, the relevant literature was comprehensively discussed to discuss the abundance, characteristics, distribution and influencing factors of MPs in farmland soil. The conclusions are as follows: (1) The highest and lowest MPs abundance were in marginal tropical humid and plateau temperate semi-arid regions, accounting for 7579 n/kg and 48 n/kg, respectively. (2) The main shapes of MPs in farmland soil are fragment/flake and fiber, accounting for 44.0 % and 34.4 %, respectively. The MPs are mostly transparent (21.8 %) and black (21.5 %). Among the MPs types, polyethylene (PE) and polypropylene (PP) are dominant, accounting for 26.2 % and 19.0 %, respectively. The main size of MPs in farmland soil is 0.1–0.5 mm, with average proportions was 51.4 %. (3) Compared with non-fertilizing and non-mulching, the MPs abundance in the fertilizing and mulching farmland soils increased by 170 % and 232 %, respectively. (4) In the farmland soil, the MPs abundance was significantly positive with temperature, sunshine hour, and altitude. (5) In farmland soil of China, the most commonly used MPs dispersion treatment was H2O2 solution digestion, the extracting solution commonly used for density flotation was NaCl solution, and microscopic and spectroscopic measurements were typically used measurements. The results could provide a basis for monitoring the MP abundances in farmland soil and preventing the transfer of MPs pollution in soil.

Comparative study of ensemble models of deep convolutional neural networks for crop pests classification

Pest infestations on wheat, corn, soybean, and other crops can cause substantial losses to their yield. Classification of crop pests is of considerable importance for accurate and intelligent pest control. Ensemble models can effectively improve the accuracy of crop pests classification and different ensemble models can produce different results. To study the advantages and disadvantages of ensemble models under different agricultural production environments, six basic models are trained on a D0 dataset. Then, three models with the best classification performance are selected. Finally, the ensemble models, i.e., linear ensemble named SAEnsemble and nonlinear ensemble SBPEnsemble, are designed to combine the basic models for crop pests classification. The accuracies of SAEnsemble and SBPEnsemble improved by 0.85% and 1.49% respectively compared to the basic model with the highest accuracy. Comparison of the two proposed ensemble models shows that the accuracy of SAEnsemble is lower when the pests are very similar to the background, while SBPEnsemble is more likely to make wrong predictions under the condition of huge difference between pests and background. Therefore, in practical agricultural production, choosing linear or nonlinear ensemble methods according to different situations can effectively improve the accuracy of crop pests classification.