Remote sensing technology enables broad-scale ecological studies, with birds indicating global changes and validating its data. Despite its global relevance, the integration of remote sensing in ornithological research in India remains limited and fragmented. We conducted extensive literature review on the remote sensing applications in ornithological studies in India from 1992 to 2022 to reorient ornithological research primarily focusing on the existing satellite data, telemetry and GIS tools. The objectives of the review are: (1) to provide an overview of remote-sensing data applications in avian ecological studies, (2) to report the spatiotemporal trends and patterns in the characteristics of remote sensing data (scale, resolution, data source) applied in the studies, and identify the region-specific approaches and (3) to identify the research gaps, improve the methods and propose priority research themes by outlining the future scope of its applications in avian ecology and the conservation efforts as an opportunity to strengthen avian research using remote sensing technology in rapidly changing landscapes. We systematically reviewed literature (N = 108) that related remotely sensed data to bird distribution, abundance assessment, and migration. Our review covered 132 bird species across 19 orders and 5 feeding guilds. Forest birds were more frequently studied than wetland birds, with nearly 50% of the studies focused on single species. Galliformes were the most represented while Falconiformes, Columbiformes, and Ciconiiformes were the least. Studies largely focused on habitat suitability using avian occurrence data (51%) while avian disease outbreak, genetics, and nest site data (each 2%) had least contribution. Our review suggests the need to bridge ecological knowledge gaps using remote sensing and GIS tools through interdisciplinary collaboration among ornithologists, remote sensing experts, and scientists from other allied fields as this can improve data-sharing policies and develop innovative user-friendly products for applications in avian research. Additionally, community participation through citizen science can generate valuable crowd-sourced data for bird conservation and ecosystem management.

Abstract - This paper presents Travel AI, an end-to-end intelligent travel platform designed to revolutionize the travel planning and management ecosystem. The system integrates multiple advanced machine learning models within a modern full-stack architecture to automate and personalize the entire travel lifecycle. Core innovations include a multi-algorithm itinerary planner that synergistically combines K-Means clustering for point-of-interest grouping with genetic algorithms for optimal activity sequencing under user constraints. A hybrid recommendation engine employs both collaborative and content-based filtering, enhanced by deep learning models for analyzing accommodation imagery and reviews. Predictive analytics are delivered through Long Short-Term Memory (LSTM) networks for dynamic flight and hotel price forecasting, empowering users with cost-efficient booking strategies. Real-time user interaction is facilitated by a transformer-based Natural Language Processing (NLP) chatbot, providing 24/7 conversational support. Critical security is embedded via a fraud detection module using Isolation Forest for anomaly detection and Logistic Regression for transaction classification. Additionally, Convolutional Neural Network (CNN)-based landmark recognition enables visual discovery. Deployed using a ReactJS frontend, Django backend, and cloud infrastructure, Travel AI establishes itself as a comprehensive, intelligent travel companion that effectively addresses prevailing challenges in personalization, financial optimization, and security within the global travel industry. Key Words: Artificial Intelligence, Machine Learning, Travel Recommendation, Itinerary Planning, Price Forecasting, LSTM, Chatbot, Fraud Detection, ReactJS, Django, Cloud Computing.

Watershed geomorphology acts as a physical template upon which a range of biotic and abiotic processes operate, and a thorough geomorphic understanding of watersheds provides a key foundation of knowledge upon which a full range of environmental management decisions can be developed. While British Columbia (BC) is home to a long history of study and significant advances in the science of geomorphology and related fields, this knowledge has not been incorporated into ecosystem management or planning to its full potential. We argue that a geomorphologically-informed approach to river management needs to be developed and adopted to support solutions to a range of river management challenges. The perspectives presented in this paper originate from a 3-day workshop that was attended by experts in geomorphology, ecology, hydrology, and habitat restoration in September 2024 in Campbell River, Vancouver Island. This workshop allowed practitioners to express their views on positive and negative aspects of current river management practice and how the River Styles Framework could be used to operationalize a geomorphologically-informed approach to river management in BC.

Predicting marine animal movements from satellite tracking data remains challenging, limiting conservation and ecosystem management efforts. To address this, we trained the Temporal Fusion Transformer (TFT) neural network on tracking data from 434 southern elephant seals to forecast locations and fill data gaps (imputation) within 7-day windows. Compared to state-space models, TFT reduced location errors by 15% and produced more efficient prediction regions, identifying where seals were likely to be found while using less area: a fivefold reduction for forecasting and 30-40% reduction for imputation. The model performed best near the continental shelf and at low-to-moderate movement speeds, with bathymetry, water temperature and current direction being the most influential environmental factors affecting the model output. When applied to new geographic regions not represented in the training dataset, model performance declined by approximately 30% across most evaluation metrics, indicating challenges in transferring learned patterns to unfamiliar environments. Our findings show that deep learning is a promising tool for analyzing large, sparse tracking datasets. The enhanced predictive capabilities have potential for dynamic conservation measures, such as forecasting the spatial evolution of animals to minimize conflicts with human activities and environmental disturbances.

Reservoir ecosystems are under increasing pressure from ecological degradation and land-use conflict. This study proposes an integrated framework combining ecosystem service (ES) evaluation, trade-off analysis, and service bundling to support ecological functional zoning in the Danjiangkou Reservoir Area, a key water source in China. Six ESs were assessed from 2010 to 2023, revealing spatial heterogeneity and rising trends in water- and climate-related services. Over time, ES interactions shifted from trade-off dominance to synergy, indicating improved ecological coordination. Cluster analysis identified two ES bundles in 2010 and four stable bundles after 2015. Based on bundle composition and inter-service dynamics, four functional zones were delineated: regulation protection, restoration priority, agricultural production core, and water yield-conservation core. These zones reflect dominant ES structures and offer a practical basis for differentiated, adaptive ecosystem management in reservoir regions. The proposed method provides a transferable tool for ecological zoning in data-rich reservoir watersheds, offering new insights for adaptive ecosystem governance.

The accurate prediction of dissolved oxygen (DO) concentration in rivers is very important for the management of aquatic ecosystems, However, the hybrid model of ' modal decomposition + prediction ' for predicting the nonlinear change of dissolved oxygen in rivers is still insufficient. In this paper, a frequency division prediction framework based on the optimal ensemble of Complementary Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) is proposed. The dissolved oxygen sequence was decomposed into multiple components by CEEMDAN, and the long short-term memory network (LSTM), support vector regression (SVR) and multi-layer perceptron (MLP) models were constructed to predict each component independently. An innovative grid search algorithm with constraints is constructed, and the advantages of each model are complemented by dynamic combination. The optimal ensemble scheme is obtained with the goal of minimizing the mean absolute error ( MAE ) of the training set. The empirical study of monitoring sections A and B in the Ganjiang River Basin shows that : in the prediction task, the prediction of the training set, the MAE of the integrated model is 18.6–35.5% lower than that of the ensemble model, the root mean square error ( RMSE ) is 22.1–22.8% lower, and the determination coefficient ( R2 ) reaches 0.954 and 0.972. In particular, the error accumulation of MAE in the 3-day prediction is 27.2–81.4% lower than that of the mixed model. This framework enables the modes of multi-component dissolved oxygen series prediction to be effectively aliasing, and provides an extensible technical path for the intelligent management of the basin.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-025-27163-x.

Introduction Efficient monitoring of marine ecosystems requires reliable and energy-efficient wireless communication to support large-scale underwater sensor networks. Cognitive radio-based Internet of Underwater Things (IoUT), particularly when integrated with non-orthogonal multiple access (NOMA), enables dynamic spectrum access but faces challenges from Rayleigh and Rician fading conditions. Spectrum sensing (SS) is therefore critical for ensuring robust and efficient spectrum utilization. Methods This work evaluates two spectrum sensing techniques for underwater cognitive radio systems: Double Match Filter (DMF) with a fixed probability of false alarm (Pfa = 0.5) and Energy Spectrum Sensing (ESS) with Pfa < 0.5. Both methods were analyzed under Rayleigh and Rician channels to reflect dynamic marine environments. Key performance metrics—probability of detection (Pd), Pfa, and bit error rate (BER)—were simulated and compared against conventional SS and Match Filter methods. Results The proposed DMF–ESS approach achieves superior detection accuracy and communication reliability, offering higher Pd for equivalent Pfa levels and lower BER across a range of SNR conditions. These gains are consistent across both fading models. Discussion By enhancing detection performance and energy efficiency, the DMF–ESS framework improves the reliability and scalability of underwater IoUT networks. This supports real-time monitoring of water quality, biodiversity, and pollution, contributing to more effective marine ecosystem management.

BackgroundMarine ecosystems worldwide face extreme stress from human activities, with the North Sea being particularly affected and experiencing altered processes. To assess anthropogenic drivers for sustainable management, the Millenium Ecosystem Assessment (MEA) and the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES) distinguished five main anthropogenic drivers: direct exploitation of fish and seafood, sea use change, human-driven climate change, pollution, and invasive alien species. However, evidence of the drivers’ relevance and their potential effects on species and the environment over time remains scarce. This systematic map provides knowledge on the five main anthropogenic drivers in the North Sea from 1945 to 2020 and identifies potential knowledge gaps in terms of management implications.MethodsTo identify relevant articles we used our published systematic map protocol. We conducted systematic searches of academic and grey literature in English, German, and French in online databases (Web of Science, Scopus, PubMed, AquaDocs). The search followed a Population-Exposure-Comparison-Outcome framework and included the period January 1945 to December 2020. A total of 22,511 articles were deduplicated and screened by title and abstract, the remaining 5795 were screened full-text to provide a widely integrated evidence base. A set of 3356 articles were retained following eligibility criteria and were included in the final database. We extracted information on drivers in detail and their effects on study populations within different areas in the North Sea. Knowledge clusters and gaps were identified from the scientific effort and are synthesized narratively.ResultsOut of the 3356 articles, the majority focused on pollution throughout the entire period of 75 years. Research interest has increased in climate change and biological invasion only in the most recent decades. We identified knowledge clusters in the southern North Sea, especially in ICES standard species areas 6 and 7, which has the most articles overall, mainly emphasizing pollution. Northern areas were in contrast studied the least. The effects of pollution were mainly linked to changes in chemical water properties and to contamination levels for benthos and fish. The other drivers were rather associated with changes in biomass or abundance, with a strong focus on fish and benthos populations. A key knowledge gap was on the effects of global change, herein defined as simultaneous assessment of all five drivers, at different organizational levels and therein on different populations.ConclusionsThis systematic map reveals substantial peer-reviewed evidence on the five main anthropogenic drivers in the North Sea. The map uncovers a strong increase in research interest regarding these drivers over the years, with a strong focus towards pollution and southern North Sea areas. Despite the increasing importance of climate change effects, this map highlights limited research effort on it. As ecosystem management nowadays strives for sustainable use of marine systems, it is more important than ever to understand linkages between drivers, potential cumulative effects and possible repercussions. The map revealed a strong knowledge gap regarding these linkages due to global change. On this basis, further systematic reviews can acknowledge these gaps, identifying the drivers’ impacts and their quick evolvement to support management decision-making at various governance levels.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13750-025-00377-2.

Identifying potential service flow and actual service flow: Source-flow-sink-use framework and case study.

Soil moisture variability of typical mixed forests in the north-south climatic transitional zone of China.

Evaluation of cadmium toxicity in Penaeus monodon using ecotoxicological tools: A foundational step for deriving seawater quality criteria.

Global meta-analysis of river bacterial communities and responses to human activities.

A new green protocol for selecting translocation sites for coastal cliff endemic plants: a case study from southern Italy.

High-throughput suspect screening of priority hormone-disrupting chemicals in megacity aquatic ecosystems using a domain-specific database framework.

Community interactions drive phytoplankton regime shifts in interconnected water bodies: Insights from space for time approach.

Characterization of multi-scale urban habitat wildness: Integration of the rewilding theory into novel urban ecosystem restoration and management

Ship-based eDNA tracking unveils early dispersal patterns and microecological dynamics of Ulva prolifera micropropagules in yellow sea green tide outbreaks.

Predicting coral reef habitat distribution in the South China Sea under climate change using MaxEnt modeling.

ABSTRACT Aim Quantifying the speed of invasive species range expansion and the mechanisms behind it is a key management goal that also informs ecological theories of spread. Byers et al. found that time since first global introduction (TSI) was the strongest predictor of non‐native range size of coastal marine invertebrates relative to environmental and species traits, and species were predicted to expand on average 400 km/decade along coastlines. Here, one decade later using the same 138 marine invertebrates, we repeat that analysis and test the prediction that the average invader expansion speed was 400 km from 2014 to 2024. Methods Using the Global Biodiversity Information Facility (GBIF), we downloaded species occurrences to estimate the non‐native range size of each invader along coastlines in the northern and southern hemispheres. We calculated expansion speed as the difference between each species' 2024 range size and its 2014 range size estimated by Byers et al. We tested (Hypothesis 1) TSI is still the most important variable for explaining global invasive range size; (Hypothesis 2) Mean expansion speed is 400 km/decade; (Hypothesis 3) Expansion speed is best predicted by distributional characteristics of a species' non‐native range (i.e., initial range size, initial number of coastlines occupied); and (Hypothesis 4) Expansion speeds vary across taxonomic groups. Results TSI was still the best predictor of non‐native range size (RVI = 1, β : 0.39–0.42). However, average range expansion speed since 2014 was 3000 km/decade, ~8× the previous prediction. Species that started with large ranges and distributions across multiple coastlines exhibited the fastest expansion speeds ( R 2 = 0.4), but there were no taxonomic patterns in expansion. Main Conclusions Ranges of non‐native species are not at equilibrium and are still spreading rapidly, posing a challenge for coastal ecosystem management.

Environmental gradients drive the ecological dynamics of bacterioplankton in the East China Sea based on eDNA metabarcoding.