PSMA-PET-Guided Intraprostatic Boost in Prostate SBRT (PROBE): A Phase 2 Trial.

This paper presents a detailed implementation study of a fully deployed, cloud-native, ML-based system for automated answer sheet evaluation, designed specifically for Indian higher-education institutions. Building on a prior theoretical proposal, this work documents the architectural decisions, algorithmic choices, and engineering trade-offs that arose during real-world deployment across a multi-tenant institutional hierarchy encompassing colleges, branches, classes, subjects, and students. The system accepts scanned handwritten student PDFs, rasterises each page at 3× scale, produces three preprocessing variants per page (original, strong-contrast binarised, and notebook-clean binarised), selects the highest-scoring variant, and submits it to the Google Cloud Vision API. The resulting raw OCR text passes through a four-layer correction pipeline: Unicode glyph normalisation, domain-specific OCR word-error correction, structural regex repair, and a final refinement step via Google Gemini 2.0 Flash. The corrected text is then parsed into a structured question–part hierarchy by a deterministic finite-state parser designed to tolerate OCR-induced label corruption. Scoring relies on a hybrid two-component engine that combines a multi-metric lexical scorer — incorporating the Jaccard index, bidirectional information-weighted fuzzy token coverage, character tri-gram similarity, and a containment score — with dense semantic similarity derived from Google Gemini Embedding-001 (768-dimensional vectors, cosine similarity, re-normalised to [0,1]), weighted 20% lexical and 80% semantic by default. A configurable minimum-credit heuristic ensures that semantically valid but lexically divergent answers are not inadvertently assigned zero marks. The backend is a Node.js 20/Express 4 REST API with all persistent data stored in Supabase (PostgreSQL with managed object storage), accompanied by a vanilla-JS web application offering role-separated interfaces for administrators, teachers, district-level uploaders, and students. On a validation set of 16 student scripts across three subjects, the hybrid engine achieves a Pearson r = 0.91 against teacher-assigned marks and a within-±1-mark accuracy of 81.2% — a 20- percentage-point improvement over lexical-only grading. Average end-to-end processing time is approximately 22 seconds per six-page answer sheet, enabling a 50-student cohort to be evaluated in under 20 minutes.

Valvular heart diseases are among the leading causes of cardiovascular complications worldwide, and early automated detection plays a critical role in improving patient outcomes. Traditional auscultation methods depend heavily on physician expertise and often fail to detect subtle cardiac abnormalities. Automated analysis of Phonocardiogram (PCG) signals provides a scalable and objective approach for cardiac screening. This project, CardioXAI — Improved Detection of Valvular Cardiac Abnormalities using Phonocardiogram Signals by Explainable AI, develops a hybrid deep learning framework that analyzes PCG recordings and provides interpretable diagnostic insights. The system classifies recordings into five conditions — Normal, Aortic Stenosis (AS), Mitral Regurgitation (MR), Mitral Stenosis (MS), and Mitral Valve Prolapse (MVP) — using a Dual-Branch EfficientNetB0 architecture that fuses 3-channel RGB spectrogram features with 19 handcrafted acoustic features. The system integrates two Explainable AI methods: Grad-CAM for spatial time-frequency localization on spectrograms, and SHAP for feature-level attribution. Beyond detection, the system provides cardiac phase localization (S1, Systole, S2, Diastole), an anatomical heart valve diagram highlighting the affected valve, severity grading, and automated clinical report generation. A novel cross-dataset validation experiment on the PhysioNet 2016 Heart Sound Challenge dataset quantifies domain shift using Jaccard similarity and KS statistics, providing evidence of model generalization across different recording devices.

Pairwise sequence comparisons are at the center of molecular evolutionary analyses. However, viral pairwise comparisons are challenging because extreme mutation rates and evolutionary pressure cause genomes to diverge rapidly, limiting detectable sequence similarity to fewer than 3% of virus pairs. To overcome these limitations, we compared viruses based on structural similarity, using predicted protein structures from ColabFold and Foldseek to define protein fold clusters. We represented each virus genome by its protein structural content. Pairwise similarities between viruses were then quantified using the Jaccard index based on the presence or absence of protein fold clusters. Using a recently established viral protein fold database, we compared all pairs of eukaryotic viruses in RefSeq. This approach increased the proportion of comparable viral genome pairs from 2.4% to 16.5%. Using this protein-fold representation of viruses, we were able to accurately predict viral families with an average sensitivity of 85.9%. Investigation of viral families showing limited sensitivity with this approach uncovered a laterally transferred structural cluster (Rep/NS1) broadly shared across diverse viral families and found in the avian lineage of adenoviruses. Sequence homology suggests that this Rep was acquired from Parvoviridae, but the protein is mutant in the ATPase active site, indicating possible exaptation towards a purely DNA binding function. In Gammapapillomaviruses, several E4 clusters were associated with human tropism. In summary, by representing viruses with structural protein clusters, we can classify highly divergent viruses, trace lateral gene transfer, and uncover features associated with viral host range.

Optical coherence tomography angiography (OCTA) is an emerging non-invasive ophthalmic imaging modality. The morphology and contour of the foveal avascular zone (FAZ) are critical biomarkers for the diagnosis of various ophthalmic and systemic diseases; therefore, achieving its accurate segmentation is of substantial clinical significance. To address challenges such as low contrast, indistinct boundaries, and structural confusion with the surrounding retinal vasculature in OCTA images, this paper proposes a multi-scale dilated convolution and dual attention network (MSAD-Net). By integrating a multi-scale dilated convolution module (MSDM) to capture extensive multi-scale contextual information and employing a dual attention module (DAM) to integrate complementary channel and spatial features, thereby synergistically boosting the feature representation of key regions. Experimental results demonstrated that the model achieved superior performance and robust generalization across multiple evaluation metrics, including the Dice similarity coefficient, Jaccard index, precision, and recall-on two public datasets. These findings confirm the robustness of MSAD-Net in the fine segmentation of the FAZ, providing high-precision technical support for the clinical quantitative analysis of ophthalmic diseases.

Abstract This study presents a methodological investigation into the performance–efficiency trade-off of using classical feature-based models instead of graph neural networks (GNNs) for link prediction in sparse social networks. The study aims to systematically evaluate the effectiveness of engineered topological features compared to machine learning approaches in the context of GitHub developer collaborations. Using the MUSAE GitHub dataset (37,000 nodes, 289,000 links), we compare traditional machine learning models such as Logistic Regression, Random Forest, and LightGBM with modern GNN architectures such as Graph Convolutional Networks, GraphSAGE, and Graph Attention Networks. Our key finding is that, especially on sparse graphs, the LightGBM model, using rigorously engineered features (Common Neighbours, Jaccard Similarity, Adamic-Adar, Preferential Attachment, Node2Vec similarity), consistently outperforms standard GNN implementations (e.g., 99.3% accuracy and 0.9996 ROC-AUC in the ML community). These results challenge the tendency to automatically favour complex GNNs and provide powerful methodological insight that feature-based learning for sparse networks can deliver both high performance and computational efficiency. The main contribution of this work is to provide a rigorous and data-driven guide for model selection in graph-based learning and to challenge the automatic preference for GNNs in sparse networks. We also implemented a recommendation system prototype that serves as a practical demonstration of the methodological insights obtained.

Medicinal plants are central to primary healthcare and cultural identity, especially in remote areas with limited access to modern health services. Ethnomedicinal knowledge in the study area has not yet been documented. This study aimed to document indigenous medicinal plant knowledge, assess sociodemographic influences on its distribution, and compare plant use patterns with previous Ethiopian studies to identify novel insights. Primary ethnobotanical data on medicinal plant use and indigenous knowledge were collected between June and September 2023 from 364 informants using semi-structured interviews, focus group discussions, and guided field walks. Demographic information and plant-use details (including plant species, parts used, preparation methods, and therapeutic applications) were systematically recorded, and the data were analyzed quantitatively using t-tests, ANOVA, informant consensus factor (ICF), fidelity level (FL), Jaccard similarity index (JSI), and Rahman similarity index (RSI). A total of 102 medicinal plant species belonging to 48 families were documented. Fabaceae (8%), Asteraceae (7%), and Solanaceae (6%) were the most represented families. Herbs constituted the dominant growth form (36.3%), and leaves were the most frequently used plant part (37.4%). Oral administration (56.2%) was the primary route of remedy application, whereas grinding and crushing were the most commonly employed preparation methods. ICF values were highest for external injuries (0.89) and neurological disorders (0.86). Multipurpose species, such as Cordia africana and Olea europaea, are under high pressure due to agricultural expansion and wildfires, which represent major anthropogenic threats. Medicinal plant knowledge differed across sociodemographic factors, with key informants, men, older participants, and illiterate informants reporting higher numbers of species than other groups (p ≤ 0.001, t-tests and ANOVA). Cross-cultural comparison demonstrated moderate to low similarity with other Ethiopian studies (JSI: 2.55-48.39%; RSI: 0.64-24.27%). The WBANP and surrounding districts harbor rich medicinal plant diversity and indigenous knowledge for treating human ailments. However, anthropogenic pressures threaten these resources and their cultural heritage. Future research should prioritize community-based conservation, pharmacological validation, and phytochemical studies of culturally and therapeutically important species to support evidence-based healthcare integration.

Freshwater fish biodiversity is undergoing a rapid decline worldwide due to habitat fragmentation, climate change, invasive species, overexploitation and pollution. Yet, for many regions, including biodiversity-rich but underrepresented areas such as Türkiye, large-scale assessments of species diversity and distribution remain limited. Environmental DNA (eDNA) metabarcoding offers a powerful, non-invasive and scalable tool for biodiversity assessment, capable of detecting both common and cryptic taxa. Here, we present the first nationwide, basin-scale eDNA metabarcoding survey of freshwater fishes across Türkiye's major river systems, spanning a transect of more than 6000 km and encompassing 29 sampling sites across seven principal basins. Triplicate water samples were analysed using high-throughput sequencing of the mitochondrial 12S ribosomal RNA (rRNA) gene, generating 351,392 high-quality reads (after filtration) and detecting 52 fish species, including native, invasive and several taxa previously unrecorded in their respective basins. The Eastern Mediterranean basin exhibited the highest species richness (28 species) and diversity (H' = 2.35; 1 - D = 0.88), whereas beta-diversity analyses revealed marked spatial structuring among basins (Jaccard similarity 0.17-0.58) and a clear biogeographic separation in principal co-ordinates analysis (PCoA) ordination. Importantly, eDNA metabarcoding uncovered cryptic and low-abundance taxa not detected by conventional surveys, demonstrating its complementary value and sensitivity. By providing the first comprehensive molecular baseline of Turkish freshwater ichthyofauna, this study illustrates the potential of large-scale eDNA approaches to transform biodiversity monitoring, inform conservation strategies and support management decisions in freshwater ecosystems facing accelerating environmental change.

Medicinal animals are integral to traditional healthcare worldwide, yet ethnozoological knowledge in many Ethiopian regions remains under-documented. This study documented the diversity of medicinal animals, associated cultural knowledge, preparation methods, and conservation implications among communities in Andracha District, Ethiopia. A cross-sectional ethnobiological survey was conducted from January to August 2025 using semi-structured interviews, focus group discussions (FGDs), and field observations. Sixty-five informants participated, including 25 key informants purposively selected and 40 general participants randomly chosen. FGDs refined data collection instruments and validated traditional knowledge. Quantitative indices Relative Frequency of Citation (RFC), Informant Consensus Factor (ICF), Fidelity Level (FL), and Zoological Ethnoknowledge Index (ZEI) summarized knowledge patterns, species importance, and cultural reliability. Jaccard Similarity Index (JSI) and Rahman's Similarity Index (RSI) assessed cross-community similarity. Statistical analyses, including t-tests, ANOVA, and linear regression, explored variations across gender, age, education, and experience using R software. A total of 52 medicinal animal species were documented in Andracha District. Bos taurus, Apis mellifera, and Halictus scabiosae exhibited the highest RFC values. ICF ranged from 0.60 to 0.97, with the strongest agreement observed for respiratory and gastrointestinal ailments. The highest FL was recorded for Capra aegagrus hircus (FL = 83.3%), followed by Apis mellifera (FL = 77.8%). Knowledge of medicinal animals varied significantly with age, gender, literacy, and healer experience (P < 0.05). Major threats to medicinal fauna included agricultural expansion, overgrazing, and overexploitation, while indigenous conservation practices such as sacred forest protection and community stewardship were actively maintained. Medicinal animals are crucial for healthcare, cultural identity, and livelihoods in Andracha District. Environmental pressures and generational gaps threaten both species and associated knowledge. Strengthening community-based conservation, sustainable harvesting, and intergenerational knowledge transmission is essential to safeguard this biocultural heritage. The study documents traditional practices but does not endorse ingestion or application of potentially hazardous animal products.

Wild edible plants (WEPs) play a crucial role in rural livelihoods, food security, and cultural heritage in Ethiopia. Despite their importance, comprehensive documentation of WEP diversity, utilization, and associated indigenous knowledge in Tegedie District remains limited. This study aimed to identify WEP species, evaluate their uses, assess knowledge distribution among community members, and identify threats to their sustainability. Ethnobotanical data were collected from 144 purposively selected informants across eight sites using semi-structured interviews, focus group discussions, and guided field observations. Quantitative indices, including the Relative Frequency of Citation (RFC) and Jaccard Similarity Index (JSI), were employed to assess species importance, cultural significance, and similarity with other Ethiopian districts. Preference ranking and direct matrix ranking were used to evaluate species preference, multifunctionality, and perceived threats. Data analysis was conducted using R software, and species identifications were cross-checked against the IUCN Red List. A total of 52 WEP species were documented, comprising trees, shrubs, herbs, and climbers. Fruit-bearing trees and shrubs were the most frequently cited, highly preferred, and culturally significant. The top-ranked WEPs included Cordia africana Lam., Syzygium guineense (Willd.) DC., Diospyros mespiliformis Hochst. ex A.DC., and Tamarindus indica L. Knowledge of WEPs was significantly higher among men, older individuals, illiterate participants, and key informants (P < 0.005), reflecting the cumulative and experiential nature of indigenous knowledge. Major threats to WEPs included agricultural expansion, overgrazing, firewood collection, charcoal production, and habitat degradation. WEPs were primarily harvested from forests, grazing lands, and uncultivated areas, with knowledge transferred orally and through hands-on participation in harvesting. WEPs in Tegedie District are vital for dietary diversity, food security, and cultural practices, but they face significant anthropogenic and environmental pressures. Conservation strategies, including habitat protection, sustainable harvesting, ex situ cultivation, and systematic documentation of indigenous knowledge, are urgently needed. Promoting intergenerational knowledge transfer is essential to ensure the continued availability and sustainable use of these valuable plant resources.

The widespread and inappropriate use of antibiotics in both human and veterinary medicine has accelerated the emergence and dissemination of antibiotic resistance genes (ARGs) in various environments. Companion animals, due to their close and prolonged interactions with humans, have increasingly been recognized as potential reservoirs and transmitters of ARGs. However, the extent remains largely unclear to which companion animals influence the diversity and distribution of ARGs in humans. Understanding these interactions is essential for assessing environmental pathways of antibiotic resistance transmission and for developing effective mitigation strategies within the One Health framework. We examined the profiles of ARGs and gut microbial communities among three groups: companion animals, pet owners, and non-pet owners. Quantitative polymerase chain reaction (qPCR) assays were applied to determine the abundance and diversity of representative ARGs, while 16S rRNA gene sequencing was used to characterize the composition and structure of microbial communities. Comparative and correlation analyses were conducted to evaluate the relationships between ARG distribution patterns and microbial community profiles across different host groups. Companion animals were found to possess the highest total abundance of ARGs (8.46 × 101⁰ copies/μL), while humans exhibited greater gut microbial diversity. ARGs ermB and tetQ displayed relatively high abundance in all three groups. In addition, intI-1 was significantly more abundant in pet owners than in non-pet owners. ARG profile of pet owners showed more similarity to that of their pets, assessed by the Jaccard similarity index. Age was associated with a limited subset of ARGs: sul2 and tetW decreased with age in companion animals, whereas aph(3'), cmlA, fexA and qnrS increased with age in humans. Notably, high correlation (r = -0.69/0.77) of oqxA-Megasphaera was identified, with negative correlation in animals and positive in pet owners, suggesting oqxA could be a potential key hub for ARGs dissemination. Our findings show that pet owners exhibit similar ARG profiles to those in companion animals, suggesting pet ownership may drive convergence in profiles of ARGs. Moreover, these findings provides evidence of potential resistome overlap at the human-animal interface and highlight the need to incorporate companion animals into antimicrobial control programs under a One Health framework.

Low-complexity regions (LCRs) are compositionally biased segments of proteins that play critical roles in molecular recognition, structural flexibility, and phase separation. Yet, their accurate detection remains challenging due to methodological variability among computational tools. In this study, we conducted a comprehensive benchmarking of eight widely used LCR detection methods (under multiple parameter settings) across the Homo sapiens proteome. A modular computational framework was developed to systematically compare LCR characteristics, including residue-centric analyses such as length distribution and coverage percentage. Protein-centric analyses included compositional bias, amino acid composition, and Shannon entropy. Consensus analyses revealed that regions detected by multiple tools were typically longer, more repetitive, and compositionally purer, suggesting stronger structural or functional relevance. Jaccard similarity matrices revealed distinct clustering patterns among algorithms based on shared detection principles. Additionally, entropy and purity analyses highlighted fundamental differences in sequence complexity captured by each tool. Together, these results provide a unified, reproducible framework for evaluating LCR detection performance and offer practical guidelines for reliable annotation of low-complexity regions in proteome-scale studies.

We investigated the diet of the invasive black rat ( Rattus rattus ) and its overlap with co-occurring small native mammals in protected areas of the temperate forests of southern Chile. Our study was conducted during three consecutive winters between 2022 and 2025. We collected 165 fecal samples that were pooled together by location into 26 pools to describe the diet breadth of the black rat. For metabarcoding analysis, we analyzed pooled extracts aggregated by locality × sampling period (up to six 6 pellets per pool; 21 black rat pools, four small native-mammal pools and one for the only marsupial species present in the study area, Dromiciops gliroides ), using a multi-marker strategy (trnL, COI, 16S). Results are interpreted as pool-level trophic overlap and potential interference rather than direct evidence of individual-level competition. Dietary overlap was quantified using Pianka’s index and Jaccard similarity. Rattus rattus exhibited a broad omnivorous diet spanning 37 plant families, 9 arthropod orders, and eight vertebrate families, including native rodents and D. gliroides . Small native mammals displayed narrower niches, with dietary overlap being high for arthropods (Pianka = 0.835), moderate for plants at higher taxonomic level (family level) (Pianka = 0.40), and low for plant species (Jaccard = 0.11). Substantial inter-individual variation indicated that some black rats relied heavily on anthropogenic subsidies (processed foods, exotic plants, human DNA), whereas other individual black rats overlapped directly with native taxa. Black rats demonstrated the ability to exploit both native and anthropogenic resources, resulting in trophic interference with native small mammals. This overlap increases the risk of competitive pressure and predation of small native mammals in globally significant southern temperate rainforests.

Fusarium oxysporum f. sp. lycopersici (FOL), the pathogen causing tomato wilt, is a major constraint to tomato production in India. This study evaluated virulence variability and genetic diversity among FOL isolates from major tomato-growing districts. Eighteen isolates proved pathogenicity tests on the susceptible cultivar PKM-1, where 67% showed high virulence, 22% were moderately virulent, and 11% were low virulent. Molecular amplification of high virulence isolates with ITS1 and ITS4 primers yielded an approximately 550 bp product in all isolates, confirming their identity as F. oxysporum. Genetic diversity was analysed using ITS-RFLP with five restriction enzymes, generating polymorphic fragments of 100–600 bp. UPGMA clustering based on Jaccard similarity coefficients of 0.83–0.98 separated isolates into two major clusters with distinct subclusters, indicating substantial genetic heterogeneity. These findings provide valuable baseline data for regional FOL management strategies.

To explore a deep learning (DL) model for determining the relationship between the maxillary sinus (MS) and maxillary posterior teeth (MPT) based on cone beam computed tomography (CBCT) images and measuring the distance automatically between the MS and MPT using a 3D point cloud algorithm. A CBCT dataset containing 88 maxillary sinuses (MSs) and 352 maxillary posterior teeth (MPT) was annotated, and the MS-MPT distances were measured by clinicians as the ground truth. A segmentation model for MSs and MPT in CBCT images based on the U-Net convolutional block attention (CBAM) architecture was trained and assessed using a 3-fold cross-validation strategy. Then, calibrated point clouds were reconstructed using segmented anatomical structure data, and the Euclidean distances between the MS and MPT were measured; the minimum distance was identified as the MS-MPT distance. The performance of the model in terms of segmentation and distance measurement was evaluated, and the results were compared with the ground truth. Our segmentation model achieved a mean Dice similarity coefficient (DSC) of 0.959 and a mean Jaccard coefficient of 0.922 for MSs and a mean DSC of 0.913 and a mean Jaccard coefficient of 0.851 for MPT. The MS-MPT distances determined by clinicians and the 3D point cloud method demonstrated strong consistency (ϒ > 0.993, p < 0.01). In terms of the model and clinicians, the mean negative signed error was 0.63 mm (95% CI, 0.59-0.66 mm), and the successful detection rate (SDR) for the root apex of MPT reached 70.3% at the 1 mm threshold. In this study, an automated framework that combines deep learning-driven segmentation and three-dimensional point cloud analysis was developed to quantify the relationship between the maxillary sinus and maxillary posterior teeth and achieved reliable detection accuracy across diverse anatomical variations in CBCT scans.

A novel consistency-guided and pseudo labeled training for weakly-semi-supervised medical image segmentation.

Pneumonia is a leading cause of morbidity and mortality in children, with bacterial pathogens being important etiologic agents. Most microbiome studies in pneumonia use technologies with limited taxonomical resolution and few include lung aspirate or blood samples. In this study, we assessed the microbial communities of the nasopharynx, nasopharynx/oropharynx, induced sputum, lung aspirate and blood, and recovered metagenome-assembled genomes from the same sites using shotgun metagenomics sequencing of samples from children with severe and very severe pneumonia in The Gambia. Our data show that Proteobacteria and Firmicutes were the most common phyla across the body sites, and this was largely driven by S. pneumoniae, H. influenzae/aegyptius and M. catarrhalis. Furthermore, we observed species overlap of blood and respiratory samples with average Jaccard similarity index values ranging from 34% to 58%. We recovered 60 medium and 35 high-quality MAGs in these niches including 11 S. pneumoniae , 10 H. influenzae strains and a limosilactobacillus with less than 95% Average Nucleotide Identity to any known species in GTDB-TK. We also showed that the resistomes in our MAGs were highly species specific with more than 70% of the detected AMR genes found exclusively in a single species.

The future is new: Historical versus contemporary medicinal plant knowledge in Jamaica.

Detecting and segmenting cells or tumors in histopathological images is a very challenging task. This study presents a novel technique for detecting tumors and segmenting cells in histopathological images. The EBHI-SEG dataset contains 5170 images of six types of tumor differentiation stages and the corresponding ground truth images. Normalization was applied using a specific target image. Then, the Gaussian blur technique was applied to reduce the noise in the original image. The Otsu's thresholding method was applied to obtain binary images, followed by morphological operations. The results obtained were evaluated using the Jaccard index, Intersection Over Union (IOU), Precision, Recall, F1-score, and Structural Similarity Index (SSIM), showing satisfactory results. A comparison with six other well-known methods showed that the proposed approach provides promising and sufficient results.

Long-term eutrophication remains one of the major pressures affecting freshwater lake ecosystems in Europe. Lake Durowskie (north-western Poland) has experienced sustained nutrient loading from agricultural runoff, urban activities, and hydrological connections with upstream lakes, resulting in a hypertrophic state. Since 2009, restoration measures combined with systematic ecological monitoring have been implemented to improve the lake’s ecological condition. This study evaluates long-term phytoplankton and periphyton dynamics as indicators of ecological recovery in Lake Durowskie during 2008–2025. Samples were collected from multiple lake and inflow sites and analysed in terms of taxonomic composition, abundance, biomass, and ecological characteristics. Ecological status was assessed using biological indices, including the Shannon–Wiener diversity index (H′), Pielou evenness index (E), Jaccard similarity index, Nygaard mixed trophic index, and the diatom index (DI). The results indicate relatively high phytoplankton diversity and evenness across the lake. Chlorophyta and Bacillariophyceae contributed most to species richness, whereas total biomass was largely influenced by dinoflagellates, particularly Ceratium hirundinella and Peridiniopsis berolinense. Long-term observations revealed increasing algal species richness since 2022 and low Jaccard similarity values (17–25%), indicating significant community restructuring. Although the Nygaard index consistently classified the lake as hypertrophic, improvements in the periphyton diatom index at several sites suggest gradual ecological improvement. Overall, the results indicate a transitional recovery phase under persistent eutrophic pressure and confirm the value of phytoplankton and periphyton as indicators for long-term monitoring of hypertrophic lake ecosystems.