Stream Environment Research Articles

Contribution of Seasonal Streams to the Conservation of Native and Migratory Birds in a Coastal Region Undergoing Urbanization

Urbanization has drastically transformed natural landscapes, threatening biodiversity by reducing species richness and abundance in cities. In this context, seasonal streams serve as refuge areas for various bird species. To determine the role of seasonal streams in conserving native terrestrial avifauna in Algarrobo, an expanding urban environment, we compared bird diversity in ephemeral streams (n = 18) and residential areas (n = 18). Bird counts were conducted during summer and winter. We utilized a diversity index and Generalized Linear Mixed Models (GLMMs) to assess the richness and abundance of native terrestrial birds, as well as the abundance of six migratory birds. Additionally, species accumulation curves verified whether most present species had been adequately sampled. Our results revealed that seasonal streams harbored a greater richness of native bird species than residential areas. Endemic species were also recorded in these streams. The cumulative number of bird species was higher in the seasonal stream environment compared to residential areas. During winter, a higher abundance of migratory birds (Green-backed Firecrown and Gray-hooded Sierra Finch) was observed in seasonal streams. These findings suggest that seasonal streams play a crucial role in conserving native and migratory birds. Thus, protecting and managing these habitats is essential for maintaining avian biodiversity in urbanized areas.

Familiar Yet New: How Design-Driven Innovation and Brand Image Affect Green Agricultural Product Purchase Intentions in the Live Streaming Environment

With the rapid development of live streaming e-commerce, green agricultural products have become an important consumer category. However, sales still face challenges such as weak brand effects, content homogeneity, and the lack of professional hosts. Research shows that various factors influence consumers’ purchase intentions, with design-driven attributes and brand image playing crucial roles. However, their impact in the context of green agricultural product live streaming remains underexplored. This study, based on the S-O-R theory, investigates the factors that stimulate consumer purchase intentions for green agricultural products and reveals the influence of design-driven attributes on purchase intentions. A total of 472 valid responses were collected through a questionnaire. The results indicate that social presence and brand image have a positive impact on purchase intention, with green perceived value and emotional attitude acting as full mediators. However, design-driven attributes do not have a significant direct impact on purchase intention. Nevertheless, emotional attitude plays a significant mediating role between design-driven attributes and purchase intention. This study contributes to the research on consumer behavior and perceived value in live streaming environments, particularly emphasizing the importance of design-driven attributes, and provides insights for optimizing live streaming strategies and improving agricultural product design.

Damming of streams due to the construction of a highway in the Amazon rainforest favors individual trophic specialization in the fish (Bryconops giacopinii).

In Amazonian streams, damming caused by road construction changes the system's hydrological dynamics and biological communities. We tested whether the degree of specialization in fish (Bryconops giacopinii) individuals is higher in pristine stream environments with intact ecological conditions than in streams dammed due to the construction of a highway in the Amazon rainforest. To achieve this, stomach content data and stable isotopes (δ13C and δ15N) in tissues with varying isotopic incorporation rates (liver, muscle, and caudal fin) were used to assess the variation in consumption of different prey over time. The indices within-individual component (WIC)/total niche width (TNW) and individual specialization were employed to compare the degree of individual specialization between pristine and dammed streams. The condition factor and stomach repletion of sampled individuals were used to infer the intensity of intraspecific competition in the investigated streams. The species B. giacopinii, typically considered a trophic generalist, has been shown to be, in fact, a heterogeneous collection of specialist and generalist individuals. Contrary to our expectations, a higher degree of individual specialization was detected in streams dammed by the highway. In dammed streams, where intraspecific competition was more intense, individuals with narrower niches exhibited poorer body conditions than those with broader niches. This suggests that individuals adopting more restricted diets may have lower fitness, indicating that individual specialization may not necessarily be beneficial for individuals. Our results support the notion that intraspecific competition is an important mechanism underlying individual specialization in natural populations. Our results suggest that environmental characteristics (e.g., resource breadth and predictability) and competition for food resources interact in complex ways to determine the degree of individual specialization in natural populations.

Temporal–spatial correlation and graph attention-guided network for micro-expression recognition in English learning livestreams

Micro-expressions, fleeting facial movements lasting 1/25 to 1/3 of a second, offer crucial insights into genuine emotions, particularly valuable in online education settings. The rapid growth of English learning livestreams has heightened the need for accurate, real-time micro-expression recognition to enhance learner engagement and instructional effectiveness. However, existing methods need help with the subtle nature of these expressions, especially in dynamic, low-resolution streaming environments. This paper presents TSG–MER–ELL, a novel end-to-end network for micro-expression recognition in English learning livestreams, integrating temporal–spatial correlation and graph attention mechanisms. The framework addresses the unique challenges of real-time emotion analysis in online language education, where subtle facial cues are crucial in understanding learner engagement and comprehension. The temporal–spatial correlation module employs action units with spatio-temporal graph convolution to aggregate features from diverse facial regions, while transformer encoders construct long-range correlations. The graph attention module builds upon local facial areas to guide self-attention computations, yielding precise local correlation features. These global and local features are fused for the final micro-expression classification. We introduce an adaptive loss function that balances accuracy, efficiency, and relevance to linguistic context. Extensive experiments on SMIC, CASME II, and SAMM datasets, adapted for English learning scenarios, demonstrate TSG–MER–ELL’s superior performance over ten state-of-the-art baselines. The TSG–MER–ELL framework achieves top UF1 and UAR scores across all datasets, significantly improving recognition speed and accuracy. Ablation studies and visualizations of temporal–spatial features and graph attention weights provide insights into the framework’s effectiveness in capturing subtle emotional cues. TSG–MER–ELL’s robust performance in varied online learning conditions highlights its potential to enhance engagement, personalize instruction, and improve overall outcomes in virtual English language education.

Graph convolutional network for fast video summarization in compressed domain

Video summarization is the process of generating a concise and representative summary of a video by selecting its most important frames. It plays a vital role in the video streaming industry, allowing users to quickly understand the overall content of a video without watching it in its entirety. Most existing video summarization methods require fully decoding the video stream and extracting the features with a pre-trained deep learning model in the pixel domain, which is time-consuming and computationally expensive. To address this issue, this paper proposes a novel method called Graph Convolutional Network-based Compressed-domain Video Summarization (GCNCVS), which directly exploits the compressed-domain information and leverages graph convolutional network to learn temporal relationships between frames, thereby enhancing its ability to capture contextual and valuable information when generating summarized videos. To evaluate the performance of GCNCVS, we conduct experiments on two benchmark datasets, SumMe and TVSum. Experimental results demonstrate that our method outperforms existing methods, achieving an average F-score of 53.5% on the SumMe dataset and 72.3% on the TVSum dataset. Additionally, the proposed method shows Kendall's τ correlation coefficient of 0.157 and Spearman's ρ correlation coefficient of 0.205 on the TVSum dataset. Our method also significantly reduces computational time, which enhances the feasibility of video summarization in video streaming environments.

Quantifying Streambed Grain Size, Uncertainty, and Hydrobiogeochemical Parameters Using Machine Learning Model YOLO

AbstractStreambed grain sizes control river hydro‐biogeochemical (HBGC) processes and functions. However, measuring their quantities, distributions, and uncertainties is challenging due to the diversity and heterogeneity of natural streams. This work presents a photo‐driven, artificial intelligence (AI)‐enabled, and theory‐based workflow for extracting the quantities, distributions, and uncertainties of streambed grain sizes from photos. Specifically, we first trained You Only Look Once, an object detection AI, using 11,977 grain labels from 36 photos collected from nine different stream environments. We demonstrated its accuracy with a coefficient of determination of 0.98, a Nash–Sutcliffe efficiency of 0.98, and a mean absolute relative error of 6.65% in predicting the median grain size of 20 ground‐truth photos representing nine typical stream environments. The AI is then used to extract the grain size distributions and determine their characteristic grain sizes, including the 10th, 50th, 60th, and 84th percentiles, for 1,999 photos taken at 66 sites within a watershed in the Northwest US. The results indicate that the 10th, median, 60th, and 84th percentiles of the grain sizes follow log‐normal distributions, with most likely values of 2.49, 6.62, 7.68, and 10.78 cm, respectively. The average uncertainties associated with these values are 9.70%, 7.33%, 9.27%, and 11.11%, respectively. These data allow for the computation of the quantities, distributions, and uncertainties of streambed HBGC parameters, including Manning's coefficient, Darcy‐Weisbach friction factor, top layer interstitial velocity magnitude, and nitrate uptake velocity. Additionally, major sources of uncertainty in grain sizes and their impact on HBGC parameters are examined.

Research on the Influence of Anchor Influence on Consumers' Purchase Intention in Live Broadcast Room under the Background of Interest E-commerce

In the emerging shopping mode of interest e-commerce live streaming, anchors are the key factors influencing consumers' purchase intentions, and their personal characteristics and behaviors have a significant impact on consumer decision-making. Based on the stimulus-organism-response (S-O-R) model, this study constructed a structural equation model under the framework of influencer marketing theory to explore the internal mechanism between the various dimensions of anchor influence, consumer trust and identity, and its impact on consumers' purchase decision-making behavior. The results show that in different confidence intervals, the five dimensions of anchors' professionalism, interactivity, authenticity, popularity and innovation can significantly enhance consumers' trust and identity, and then significantly improve consumers' purchase intention. In addition, consumers' trust and recognition play a mediating role between the influence of anchors and consumers' purchase intentions. This study can enrich the theoretical discussion of consumer behavior decision-making in the live streaming environment of interest e-commerce, and provide a strategic reference for e-commerce anchors and interest e-commerce platforms.

Dominance and size affect foraging position choice by Arctic grayling (Thymallus arcticus).

Salmonid fishes are a widespread and economically important group of fishes that are strongly affected by anthropogenic environmental degradation. Consequently, studies on their behavior and ecology are essential for their conservation. Wild Arctic grayling (Thymallus arcticus) were observed in an unimpacted Alaskan river using underwater video observations for two consecutive summers (2015 and 2016). Foraging, aggressive behavior, and position held by individual fish were quantified. Fish held focal positions in the water column most of the time (mean = 81%), active foraging was observed for 14% of the observation periods, and antagonistic interactions occurred during 5% of the time. The analysis of intraspecific aggression revealed that aggressive interactions occurred between conspecifics of similar sizes (±100 mm), although the winners of bouts were significantly larger than the losers. Notably, initiators emerged as more successful in these bouts compared to receivers, irrespective of body length. These findings underscore the significance of conducting field studies to gain valuable insights into the behavior of fish in their natural stream environments.

ACroSS: AI-driven cross-layer adaptive streaming for short video applications

As short video applications gain popularity, researchers are exploring ways to enhance Quality of Experience (QoE) for short videos while maximizing network bandwidth efficiency. Despite the growing interest, existing Adaptive Bitrate (ABR) algorithms primarily concentrate on content prefetching strategies and often overlook the dynamic interaction between network congestion control and ABR. This interaction is especially critical for short video streaming, where network conditions can fluctuate rapidly, and user expectations for seamless playback are high. To address these challenges, we propose aCroSS, an AI-driven framework for adaptive short video streaming that jointly optimizes both the application and transport layers to enhance QoE and bandwidth utilization. The aCroSS algorithm leverages advanced machine learning techniques to adapt in real time to fluctuating network conditions and dynamic user behaviors, delivering a more robust and responsive streaming experience. Our simulation results demonstrate that aCroSS consistently outperforms existing baseline algorithms, achieving more than a 10% improvement in utility scores across various network trace datasets. This highlights the effectiveness of aCroSS in delivering superior performance in diverse streaming environments.

An Experimental Methodology for Automated Detection of Surface Turbulence Features in Tidal Stream Environments

Tidal stream environments are important areas of marine habitat for the development of marine renewable energy (MRE) sources and as foraging hotspots for megafaunal species (seabirds and marine mammals). Hydrodynamic features can promote prey availability and foraging efficiency that influences megafaunal foraging success and behaviour, with the potential for animal interactions with MRE devices. Uncrewed aerial vehicles (UAVs) offer a novel tool for the fine-scale data collection of surface turbulence features and animals, which is not possible through other techniques, to provide information on the potential environmental impacts of anthropogenic developments. However, large imagery datasets are time-consuming to manually review and analyse. This study demonstrates an experimental methodology for the automated detection of turbulence features within UAV imagery. A deep learning architecture, specifically a Faster R-CNN model, was used to autonomously detect kolk-boils within UAV imagery of a tidal stream environment. The model was trained on pre-existing, labelled images of kolk-boils that were pre-treated using a suite of image enhancement techniques based on the environmental conditions present within each image. A 75-epoch model variant provided the highest average recall and precision values; however, it appeared to be limited by sub-optimal detections of false positive values. Although further development is required, including the creation of standardised image data pools, increased model benchmarking and the advancement of tailored pre-processing techniques, this work demonstrates the viability of utilising deep learning to automate the detection of surface turbulence features within a tidal stream environment.

Microplastics and riverine macroinvertebrate communities in a multiple-stressor context: A mesocosm approach

Growing use of synthetic materials has increased the number of stressors that can degrade freshwater ecosystems. Many of these stressors are relatively new and poorly understood, such as microplastics which are now ubiquitous in freshwater systems. The effects of microplastics on freshwater biota must be investigated further in order to better manage and mitigate their impacts. Our experiment provides the first empirical evaluation of stream invertebrate community dynamics in response to microplastics of different concentrations and sizes, in combination with fine sediment, a pervasive known stressor in running waters. In a 7-week streamside experiment using 64 flow-through circular mesocosms, we investigated the effects of exposure to three simulated microplastic influxes (polyethylene microspheres at four levels between 0 and 28,800 items/event) and the addition of fine sediment (to simulate a polluted stream environment). Invertebrate drift was monitored for 48 h immediately after each microplastic influx, and benthic invertebrate communities were sampled after 28 days of microplastic and sediment manipulations. Microplastic concentration, size and fine sediment all had significant factor main effects on several invertebrate drift response metrics, whereas few microplastic main effects were seen in the benthic community. However, interactive stressor effects were common in different combinations between sediment, microplastic size and concentration, suggesting multiple-stressor relationships between microplastics and fine sediment. Microplastic ingestion was witnessed in four of 12 taxa analysed: Hydrobiosidae, Deleatidium spp., Potamopyrgus antipodarum and Archichauliodes diversus. Our findings provide insights into how microplastics affect drift and benthic community dynamics of stream invertebrates in a field-realistic experimental setting and highlight areas requiring further study. These include investigations of invertebrate drift dynamics in response to other types of microplastics, the role invertebrate size may play in determining their vulnerability to microplastic pollution, and framing more microplastic research in a field-realistic multiple-stressor context.

The Impact of Household Pollutants and Fish Assemblages on the Environment of Some Internal Streams Fed by the Euphrates River, North of Basrah Province

The Impact of Household Pollutants and Fish Assemblages on the Environment of Some Internal Streams Fed by the Euphrates River, North of Basrah Province

MEMF: Multi-entity multimodal fusion framework for sales prediction in live streaming commerce

Live streaming commerce thrives with a rich tapestry of multimodal information that intertwines with various entities, including the anchor, the commodities, and the live streaming environment. Despite the wealth of data at hand, the synthesis and analysis of this information to predict sales remains a significant challenge. This study introduces a framework for multi-entity multimodal fusion, which is characterized by the effective synthesis of multimodal data and its prioritization of entity-level fusion, thereby providing a comprehensive feature representation for improving predictive performance. In addressing the multimodal data associated with a diverse range of products, our framework improves the Transformer architecture to initially capture the intra-product modal features and subsequently integrate the inter-product features. Data experiments are conducted on a real-world dataset from Taobao Live. The framework outperforms both traditional machine learning methods and state-of-the-art multimodal fusion methods, which affirms its value as a robust decision-support tool for sales prediction, enabling more accurate pre-event predictions and strategic planning. We also examine the impact of different types of information in accurate sales prediction. It is found that harnessing a comprehensive suite of data leads to optimal performance across all evaluation metrics. Commodity-related data is primary factor in determining the prediction accuracy, followed by video data and streaming room-related data, providing insights regarding the resource allocation for collecting and analyzing multimodal data from live streaming platforms.

A new giant Jurassic lacewing larva reveals a particular aquatic habit and its significance to the palaeoecology

AbstractNeuroptera, as a small relic group of Insecta undergoing a rapid species diversification during the Mesozoic Era, is known by diverse extinct endemic lineages preserved as impression fossils and in amber. The current understanding of Mesozoic neuropterans′ diversity has mainly focused on the adults, because the contemporaneous larvae have been fairly rare especially for the Jurassic lacewings. Herein, a new giant lacewing larva, Natator giganteus gen. et sp. nov., is described from the Middle Jurassic Daohugou Beds of China. The remarkable larva is characterized by its impressively large body size, distinctively elongated cervix, and presence of swimming hairs on legs, which provide direct evidence to reveal an aquatic habit for the Jurassic lacewing larva. The morphological analysis indicates this giant larva would have probably inhabited the benthic environments of Jurassic montane rivers and streams. In addition, its morphological specialization suggests that it might have adopted an ambush predation strategy to catch its prey. The finding enhances our knowledge of the species diversity and morphological plasticity for the Jurassic lacewing larvae, and reveals that the aquatic lineages of Neuroptera exhibited dramatically structural and ecological convergence across the evolutionary process.

Dynamic Data Scaling Techniques for Streaming Machine Learning

This research delves into innovative dynamic data scaling techniques designed for streaming machine learning environments. In the realm of real-time data streams, conventional static scaling methods may encounter challenges in adapting to evolving data distributions. To overcome this hurdle, our study explores dynamic scaling approaches capable of adjusting and optimizing scaling parameters dynamically as the characteristics of incoming data shift over time. The objective is to augment the performance and adaptability of machine learning models in streaming scenarios by ensuring that the scaling process remains responsive to changing patterns in the data. Through empirical evaluations and comparative analyses, the study aims to showcase the efficacy of the proposed dynamic data scaling techniques in enhancing predictive accuracy and sustaining model relevance in dynamic and fast-paced streaming environments. This research contributes to the advancement of scalable and adaptive machine learning methodologies, particularly in applications where timely and accurate insights from streaming data are crucial.

GreenABR+: Generalized Energy-Aware Adaptive Bitrate Streaming

Adaptive bitrate (ABR) algorithms play a critical role in video streaming by making optimal bitrate decisions in dynamically changing network conditions to provide a high quality of experience (QoE) for users. However, most existing ABRs suffer from limitations such as predefined rules and incorrect assumptions about streaming parameters. They often prioritize higher bitrates and ignore the corresponding energy footprint, resulting in increased energy consumption, especially for mobile device users. Additionally, most ABR algorithms do not consider perceived quality, leading to suboptimal user experience. This paper proposes a novel ABR scheme called GreenABR+, which utilizes deep reinforcement learning to optimize energy consumption during video streaming while maintaining high user QoE. Unlike existing rule-based ABR algorithms, GreenABR+ makes no assumptions about video settings or the streaming environment. GreenABR+ model works on different video representation sets and can adapt to dynamically changing conditions in a wide range of network scenarios. Our experiments demonstrate that GreenABR+ outperforms state-of-the-art ABR algorithms by saving up to 57% in streaming energy consumption and 57% in data consumption while providing up to 25% more perceptual QoE due to up to 87% less rebuffering time and near-zero capacity violations. The generalization and dynamic adaptability make GreenABR+ a flexible solution for energy-efficient ABR optimization.

Fish conservation in streams of the agrarian Mississippi Alluvial Valley: conceptual model, management actions, and field verification

The effects of agriculture and flood control practices accrued over more than a century have impaired aquatic habitats and their fish communities in the Mississippi Alluvial Valley, the historic floodplain of the Lower Mississippi River prior to leveeing. As a first step to conservation planning and adaptive management, we developed and tested a conceptual model of how changes to this floodplain have affected stream environments and fish assemblages. The model is deliberately simple in structure because it needs to be understood by stakeholders ranging from engineers to farmers who must remain engaged to ensure effective conservation. Testing involved multivariate correlative analyses that included descriptors of land setting, water quality, and fish assemblages representing 376 stream samples taken over two decades and ranging in Strahler stream order from 1 to 8. The conceptual model was adequately corroborated by empirical data, but with unexplained variability that is not uncommon in field surveys where gear biases, temporal biases, and scale biases prevent accurate characterizations. Our conceptual model distinguishes three types of conservation actions relevant to large agricultural floodplains: reforestation of large parcels and riparian zone conservation, in-channel interventions and connectivity preservation, and flow augmentation. Complete restoration of the floodplain may not be an acceptable option to the agriculture community. However, in most cases the application of even the most basic measures can support the return of sensitive aquatic species. We suggest that together these types of conservation actions can bring improved water properties to impacted reaches, higher reach biodiversity, more intolerant species, and more rheophilic fishes.

Socio-technical systems and trust transfer in live streaming e-commerce: analyzing stickiness and purchase intentions with SEM-fsQCA

Live streaming has gained substantial significance in the e-commerce realm, attracting the attention of scholars due to its profound impact on the consumer decision-making journey. However, previous studies have not sufficiently investigated into the complex of marketing strategies through trust transfer mechanisms and socio-technical aspects, considering the dynamic nature of the live streaming e-commerce environment. Grounded on the socio-technical system theory, this research develops a research framework that centers on the exploration of interactivity, entertainment, and visualization as catalysts for trust transfer in live streaming. Trust transfer is conceptualized as advancing from cognitive trust to affective trust. Cognitive trust is proposed as a mediating factor bridging socio-technical system constructs to affective trust, ultimately shaping stickiness behavior and purchase intention in the live streaming context. To test the hypotheses, this study collected data from 682 participants in Indonesia and employed a hybrid analysis approach, combining SEM and fsQCA. The SEM results confirm that socio-technical system constructs significantly predict cognitive trust but do not directly impact affective trust. Instead, cognitive trust plays a full mediating role in transforming trust rooted in emotional bonds within the live streaming environment. Furthermore, both cognitive and affective trust exhibit a significant influence on the formation of stickiness behavior and the driving of purchase intention in the domain of live streaming e-commerce. The findings from fsQCA validate diverse configurations that shape stickiness behaviors and purchase intention, enriching marketing and communication strategies within the live streaming context. Additionally, the fsQCA configurations suggest varying conditions for high and low desired outcomes in stickiness and purchase intention, offering a comprehensive understanding of consumer behavior in live streaming e-commerce. This research makes substantial contributions to both theoretical understanding and marketing practice by providing an extensive discussion of configuration combinations that offer enhanced insights into the study's findings.

Interrelationships among litter chemistry, plant species diversity, and litter decomposition in tropical stream environments: a review

Factors that may accelerate decomposition are important for ecosystem functioning since plant litter decomposition is essential for carbon and nutrient cycling, but it is a generally slow process, which can take weeks up to years. In this context, studies have demonstrated that the chemical characteristics of litter mixtures can accelerate decomposition through several mechanisms. Tropical riparian forests are known for their high diversity of tree species, which leads to a wide array of litter types in tropical streams, each with distinct chemical properties. This underscores the key role of litter chemistry in significantly influencing the litter breakdown rate within of these streams. Here, we explore the interplay among litter chemistry, plant species diversity, and litter decomposition in tropical streams. We highlight the importance of litter physical and chemical characteristics for decomposition, as well as of the preservation of the natural floristic composition of tropical riparian forests. In this sense, more attention must be paid to the influence that the environment and phylogeny may have on the phytochemical characteristics of riparian forest plant species in different tropic biomes, and how the insertion of different exotic species interferes with the decomposition process. Furthermore, we emphasize the need for additional research into the consequences of the loss of rare plant species with unique functional characteristics to decomposition in tropical ecosystems.

Class imbalance aware drift identification model for detecting diverse attack in streaming environment

<div align="center"><span>Detecting fraudulent transactions in a streaming environment presents several challenges including the large volume of data, the need for real-time detection, and the potential for data drift. To address these challenges a robust model is needed that utilizes machine learning techniques to classify transactions in real-time. Hence, this paper proposes a model for detecting fraudulent transactions in a streaming environment using xtream gradient boost (XGBoost), cross-validation and class imbalance aware drift identification (CIADI) model. The performance of the proposed method is evaluated using datasets named credit card and Network Security Laboratory (NSL-KDD) dataset. The results demonstrate that the model can effectively detect fraudulent transactions with high accuracy, recall, and F-measure. The results show that the proposed CIADI model attained 95.63% for the credit card dataset which is higher accuracy in comparison to the generative-adversarial networks (GAN), network-anomaly-detection scheme-based on feature-representation and data-augmentation (NADS-RA) and feature-aware XGBoost (FA-XGB). Further the proposed CIADI model attained 98.5% for the NSL-KDD dataset which is higher accuracy in comparison to the NADS-RA, stacked-nonsymmetric deep-autoencoder (sNDAE) and convolutional neural-network (CNN). This study suggests that the proposed method can be an effective model for detecting fraudulent transactions in streaming environments.</span></div>