• Rural Digital Europe
• Publications close
• PLoS ONE close
• Epsilon Open Archive close

Weeds are one of the greatest threats to the growth of rice, and the loss of crops is greater in the early stage of rice growth. Traditional large-area spraying cannot selectively spray weeds and can easily cause herbicide waste and environmental pollution. To realize the transformation from large-area spraying to precision spraying in rice fields, it is necessary to quickly and efficiently detect the distribution of weeds. Benefiting from the rapid development of vision technology and deep learning, this study applies a computer vision method based on deep-learning-driven rice field weed target detection. To address the need to identify small dense targets at the rice seedling stage in paddy fields, this study propose a method for weed target detection based on YOLOX, which is composed of a CSPDarknet backbone network, a feature pyramid network (FPN) enhanced feature extraction network and a YOLO Head detector. The CSPDarknet backbone network extracts feature layers with dimensions of 80 pixels ⊆ 80 pixels, 40 pixels ⊆ 40 pixels and 20 pixels ⊆ 20 pixels. The FPN fuses the features from these three scales, and YOLO Head realizes the regression of the object classification and prediction boxes. In performance comparisons of different models, including YOLOv3, YOLOv4-tiny, YOLOv5-s, SSD and several models of the YOLOX series, namely, YOLOX-s, YOLOX-m, YOLOX-nano, and YOLOX-tiny, the results show that the YOLOX-tiny model performs best. The mAP, F1, and recall values from the YOLOX-tiny model are 0.980, 0.95, and 0.983, respectively. Meanwhile, the intermediate variable memory generated during the model calculation of YOLOX-tiny is only 259.62 MB, making it suitable for deployment in intelligent agricultural devices. However, although the YOLOX-tiny model is the best on the dataset in this paper, this is not true in general. The experimental results suggest that the method proposed in this paper can improve the model performance for the small target detection of sheltered weeds and dense weeds at the rice seedling stage in paddy fields. A weed target detection model suitable for embedded computing platforms is obtained by comparing different single-stage target detection models, thereby laying a foundation for the realization of unmanned targeted herbicide spraying performed by agricultural robots.

Producer groups are influential in Ghana’s cocoa value chain. They facilitate training, extension, education and inputs to their members. Still, there is no study on the impact of these producer groups on the technical efficiency and productivity of smallholder cocoa farmers. Using data from 217 and 199 members and non-members of cocoa producer groups, this study aimed to analyse producer groups’ impact on smallholder farmers’ technical efficiency and yield. The truncated normal distribution stochastic frontier model was adopted to estimate the farmers’ technical efficiency. Since the model showed an issue of inefficiency among the farmers, we added socio-demographic and institutional variables to account for determinants of inefficiency. Finally, we adopted an endogenous treatment regression model to analyse producer groups’ impact on the farmers’ technical efficiency by accounting for observed and unobserved biases. The study results show that farm size, labour, and capital significantly positively impact the partial elasticity of production. Age, education, use of hybrid cocoa, involvement in off-farm jobs, extension access, and producer group membership significantly affect inefficiency. The results further show that producer group membership significantly impacts technical efficiency and yield from the endogenous treatment regression model. To deal with the issues of non-participating in the producer groups, the study recommends that producer groups should be made accessible to farmers. Policymakers can promote the formation and strengthening of producer groups, leading to improved productivity and technical efficiency among cocoa farmers. This approach empowers farmers, enhances their access to resources and knowledge, and enables them to collectively address common challenges, ultimately contributing to sustainable cocoa production and better livelihoods for cocoa farming communities.

This study employs a bivariate GARCH model to examine the influence of the COVID-19 pandemic on the interactions of the commodities in the agricultural market via a connectedness network approach. Empirical results show that this pandemic alters the commodities’ roles—the activators, net transmitters, and net receivers—in the volatility and return connectedness but not for the activators in the correlation connectedness. Moreover, this pandemic enhances the interactive degree of the unidirectional negative return spillovers and the bidirectional distinct-sign volatility spillovers but doesn’t for the interactive degree of correlation. Thus, the COVID-19 pandemic, a short-term drastic event, can influence short-term interactions like volatility and return spillovers but can’t affect one long-term interaction like the correlation. Nevertheless, this pandemic raises the intensity of the correlation as well as volatility and return spillovers. The findings provide policymakers to make short- and long-term investment strategies in the agriculture market.

In 2019, the bark beetle Ips amitinus (native to central Europe) was identified in south-western Siberia at a distance exceeding 2500 km east of its previously known easternmost location in the European part of Russia. In Siberia, its invasive populations are characterised by high abundance and harmfulness. Here, I. amitinus accomplishes primary attacks on standing vital trees of Pinus sibirica with a lethal outcome. This invasion has already resulted in massive dieback in stands of pine over a large geographic territory. By, 2021, the invaded area was estimated to cover at least 31,200 km2. The objectives of this study were to investigate fungi associated with/vectored by I. amitinus in its invasive area in south-western Siberia and wood decay fungi that cause root and butt rots to P. sibirica. This led to the following conclusions: (i) DNA analysis of sixty adult beetles of Ips amitinus collected from P. sibirica in south-west Siberia revealed the presence of 143 fungal taxa; (ii) species richness was significantly higher in beetles collected from dead branches than from (more recently infested) dying branches; (iii) fungal communities were >90% dominated by yeasts, among which the most common were Nakazawaea holstii, Kuraishia molischiana, and N. ambrosiae; (iv) entomopathogenic Beauveria bassiana s.l. was the most common fungus isolated from dead/mycosed beetles of I. amitinus, followed by Lophium arboricola and four Ophiostoma spp.; and (v) Heterobasidion parviporum was the most common decay fungus detected, which was causing heart rot in stems of P. sibirica.

European forests are among the most extensively studied ecosystems in the world, yet there are still debates about their recent dynamics. We modeled the changes in tree canopy height across Europe from 2001 to 2021 using the multidecadal spectral data from the Landsat archive and calibration data from Airborne Laser Scanning (ALS) and spaceborne Global Ecosystem Dynamics Investigation (GEDI) lidars. Annual tree canopy height was modeled using regression tree ensembles and integrated with annual tree canopy removal maps to produce harmonized tree height map time series. From these time series, we derived annual tree canopy extent maps using a ≥ 5 m tree height threshold. The root-mean-square error (RMSE) for both ALS-calibrated and GEDI-calibrated tree canopy height maps was ≤4 m. The user's and producer's accuracies estimated using reference sample data are ≥94% for the tree canopy extent maps and ≥ 80% for the annual tree canopy removal maps. Analyzing the map time series, we found that the European tree canopy extent area increased by nearly 1% overall during the past two decades, with the largest increase observed in Eastern Europe, Southern Europe, and the British Isles. However, after the year 2016, the tree canopy extent in Europe declined. Some regions reduced their tree canopy extent between 2001 and 2021, with the highest reduction observed in Fennoscandia (3.5% net decrease). The continental extent of tall tree canopy forests (≥ 15 m height) decreased by 3% from 2001 to 2021. The recent decline in tree canopy extent agrees with the FAO statistics on timber harvesting intensification and with the increasing extent and severity of natural disturbances. The observed decreasing tree canopy height indicates a reduction in forest carbon storage capacity in Europe.

Non-expert users can now program robots using various end-user robot programming methods, which have widened the use of robots and lowered barriers preventing robot use by laypeople. Kinesthetic teaching is a common form of end-user robot programming, allowing users to forgo writing code by physically guiding the robot to demonstrate behaviors. Although it can be more accessible than writing code, kinesthetic teaching is difficult in practice because of users’ unfamiliarity with kinematics or limitations of robots and programming interfaces. Developing good kinesthetic demonstrations requires physical and cognitive skills, such as the ability to plan effective grasps for different task objects and constraints, to overcome programming difficulties. How to help users learn these skills remains a largely unexplored question, with users conventionally learning through self-guided practice. Our study compares how self-guided practice compares with curriculum-based training in building users’ programming proficiency. While we found no significant differences between study participants who learned through practice compared to participants who learned through our curriculum, our study reveals insights into factors contributing to end-user robot programmers’ confidence and success during programming and how learning interventions may contribute to such factors. Our work paves the way for further research on how to best structure training interventions for end-user robot programmers.

AbstractForest management operations greatly influence stream habitats. Canopy clearing and subsequent canopy development during succession, site preparation, and ditching alter the light environment, and increase sediment inputs and nutrient exports from upland and riparian soils to streams. These physicochemical changes affect aquatic biofilms and metabolic rates, and in this study, we tested their individual and combined effects. We used 12 artificial streamside channels, together with a field survey of nine streams in and around clear‐cuts, to assess the effects of shading, substrate composition, and nutrient addition on biofilm biomass and composition, as well as metabolic rates. We found that biofilm biomass and gross primary production (GPP) were light limited in channels under 70% canopy shading. Nitrate additions at this shading level only marginally increased autotrophic biomass, while the rates of respiration increased 10‐fold when carbon was added. Open (unshaded) channels had three times higher rates of GPP compared with channels with 70% shading, and autotrophic biomass was twice as high, largely caused by the colonization of filamentous green algae. These changes to biofilm biomass, composition, and GPP were caused by differences in light alone, as temperature was not affected by the shading treatment. Notably, higher rates of GPP led to no positive effect on net ecosystem production. Further, fine‐grained substrates negatively affected GPP as compared with stone substrates in the experimental channels. In the surveyed streams, the negative effects of fine‐grained substrates exceeded the positive influence of light on biofilm biomass. Altogether, our results highlight the need for riparian management that protects headwaters from unwanted stressors by focusing on preventing sediment erosion and carbon transport in clear‐cuts, while providing variable shade conditions in second‐growth forests.