Ecological Informatics最新文献

{"title":"Analysing vocal complexity in relation to sociality in orcas of British Columbia: An application of long-term computational passive acoustics","authors":"Paul Best ,&nbsp;Marion Poupard ,&nbsp;Ricard Marxer ,&nbsp;Paul Spong ,&nbsp;Helena Symonds ,&nbsp;Hervé Glotin","doi":"10.1016/j.ecoinf.2025.103211","DOIUrl":"10.1016/j.ecoinf.2025.103211","url":null,"abstract":"<div><div>Orcas are both highly social and highly vocal animals. In coastal waters of the North-Eastern Pacific Ocean, the Northern Resident orca population is well monitored, providing a great opportunity to learn about their social and communicative behaviour. Here, we report a series of acoustic analyses that lead to the empirical assessment of factors that might impact vocal complexity.</div><div>Automatically processing long-term passive acoustic data, we detected and classified calls to transcribe vocal activity. Detailed post-hoc analyses show that the detection model is imperfect, especially in detecting calls of low energy. Also, diarisation is not possible with this data and transcriptions might gather a mixture of several emitters. Taking these limitations into account, we measured communicative complexity considering the groups’ vocal production as a whole. Acoustic and visual cues also enabled the identification of specific groups with estimated numbers of individuals.</div><div>Results highlight a positive correlation between vocal and social complexity, which could be due to the mere effect of having more potential emitters. Nonetheless, this brings a first demonstration of the non-trivial link between the number of emitters and complexity in the composition of sequences. We also demonstrate significant impacts of other proximate factors such as behaviour on vocal complexity measurements, and advocate for multi-factor considerations when evaluating communicative complexity.</div><div>This work demonstrates the pertinence of joint efforts between passive acoustics, visual observations and machine learning to enhance the scale of behavioural studies and assess the validity of evolutionary hypotheses of communication systems.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"90 ","pages":"Article 103211"},"PeriodicalIF":5.8,"publicationDate":"2025-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144221559","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evolutionary polynomial modeling for interpretable drought prediction and resilient resource management","authors":"Tulio J. Francisco ,&nbsp;Bruno da Silva Macêdo ,&nbsp;Zaher Mundher Yaseen ,&nbsp;Nikolay O. Nikitin ,&nbsp;Matteo Bodini ,&nbsp;Angela Gorgoglione ,&nbsp;Camila M. Saporetti ,&nbsp;L. Goliatt","doi":"10.1016/j.ecoinf.2025.103217","DOIUrl":"10.1016/j.ecoinf.2025.103217","url":null,"abstract":"<div><div>Droughts are natural hazards that exist in nature and can have a serious impact on the environment and society, which includes water shortages, crop failures, fires and, in some cases, soil manipulation. To assess and predict droughts, various methods, such as the Standardized Precipitation Index (SPI), were designed to segregate drought trends and excess rainfall over a period ranging from 3 to 48 months. This study proposes an innovative approach to predicting drought use, the Evolutionary Polynomial Expansion with Feature Selection (EPEFS) model, a hybrid method that integrates polynomial regression with feature selection to increase accuracy and interpretability. The methodology was applied to historical precipitation data from six meteorological stations in Türkiye, covering the period from 1971 to 2016. The drought index Standardized Precipitation Index (SPI) was used as the primary indicator, with predictions made for three different time scales: SPI-3, SPI-6 and SPI-12. Furthermore, a time series cross-validation strategy was employed to ensure performance assessment. The EPEFS model obtained R<span><math><msup><mrow></mrow><mrow><mn>2</mn></mrow></msup></math></span> coefficients of 0.880, 0.903 and 0.929 for SPI-3, SPI-6 and SPI-12, respectively, surpassing the other models analyzed. Furthermore, the model presented less complexity in the generated expressions. The results suggest that the EPEFS model holds promise as a robust and interpretable tool for drought forecasting, with potential applications in early warning systems and mitigation strategies.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"90 ","pages":"Article 103217"},"PeriodicalIF":5.8,"publicationDate":"2025-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144241689","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Projecting population dynamics and range expansion of reintroduced wild boar in Scotland using agent-based modelling","authors":"Connor Lovell ,&nbsp;Terence P. Dawson ,&nbsp;J. Gareth Polhill","doi":"10.1016/j.ecoinf.2025.103261","DOIUrl":"10.1016/j.ecoinf.2025.103261","url":null,"abstract":"<div><div>The number of species reintroductions is increasing globally via both legal and illegal routes. These reintroductions can be controversial with uncertain social-ecological outcomes, particularly for unsanctioned illegal releases, which risks causing conflict between stakeholders. Despite this, current reintroduction science is focused on short-term population establishment, with little long-term modelling of reintroduced populations. In this study, we develop an agent-based model (ABM) to simulate the controversial reintroduction of wild boar in Scotland. The ABM uses probabilistic birth, death, and movement rules from the literature to stochastically simulate boar population dynamics from their initial release to 50 years in the future. Model evaluation demonstrated that the ABM behaves in predictable and explainable ways, whilst reproducing real boar behaviours and aligning with the spatial distribution of boar sightings in Scotland. Projecting the ABM 50 years into the future suggests that current boar populations are likely viable and will continue to grow and expand, with the model confirming the existence and long-term persistence of four boar populations. We conclude by commenting on the potential future uses of the ABM.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"90 ","pages":"Article 103261"},"PeriodicalIF":5.8,"publicationDate":"2025-06-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144241686","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"SegmentR: Deep learning for automated segmentation with an R interface","authors":"James D. Boyko","doi":"10.1016/j.ecoinf.2025.103259","DOIUrl":"10.1016/j.ecoinf.2025.103259","url":null,"abstract":"<div><div>The increasing digitization of biological data has generated biodiversity data at an unprecedented scale. However, extracting phenotypic information from these images poses unique challenges for biologists. Manual image segmentation is time-consuming and can be subjective, while existing automated solutions often require extensive coding experience or utilize coding languages not typically used by practicing ecologists and evolutionary biologists. Here, I present SegmentR, a user-friendly software package that leverages two state-of-the-art deep learning models – GroundinDINO and an efficient version of the Segment Anything Model (SAM). The SegmentR package provides an R-based interface, making it more accessible to biologists without coding experience. SegmentR allows users to load images, automatically segment them based on text prompts, and extract regions of interest for downstream analysis. The package includes basic visualization and data processing functions to facilitate interpretation of the results and integration with existing analytical workflows. This paper introduces SegmentR's features and demonstrates its utility through examples including isolating fish anatomy, batch processing flower images for color analysis, and segmenting museum specimens.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"90 ","pages":"Article 103259"},"PeriodicalIF":5.8,"publicationDate":"2025-06-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144231657","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A lightweight context-aware framework for toxic mushroom detection in complex ecological environments","authors":"Zhanchen Wei ,&nbsp;Jiali Wang ,&nbsp;Haohai You ,&nbsp;Ruiqing Ji ,&nbsp;Fude Wang ,&nbsp;Lei Shi ,&nbsp;Helong Yu","doi":"10.1016/j.ecoinf.2025.103256","DOIUrl":"10.1016/j.ecoinf.2025.103256","url":null,"abstract":"<div><div>The accidental proliferation of toxic mushrooms in natural ecosystems poses risks to both biodiversity and human activities in forested regions. Existing detection methods struggle with three key challenges in environmental monitoring: (1) poor discrimination of morphologically similar species in wild habitats, (2) high computational costs limiting deployment in resource-constrained field settings, and (3) performance degradation under ecological variations such as weather changes and terrain complexity. To address these challenges, we propose PM-YOLO which integrates the Contextual and Spatial Feature Calibration Network (CSFCN) and Contextual Anchor Attention (CAA) mechanisms, and is specifically designed for poisonous mushroom recognition. With the help of knowledge distillation technology, our model achieves an [email protected] with 92.64 %, which is 2.06 % higher than that of YOLOv8s. Meanwhile, the number of parameters is only 31.25 % of that of YOLOv8s (3.5 M vs. 11.2 M). Rigorous 10-fold cross-validation demonstrates its excellent robustness, with performance differences of less than 2 % across various test scenarios. PM-YOLO achieves multi-scale feature alignment through hierarchical context fusion, performs adaptive attention weighting for morphological variations, and maintains a low computational cost while significantly improving accuracy. This breakthrough enables the practical application of AI-assisted mushroom identification, effectively bridging the critical gap between academic research and field applications in the field.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"90 ","pages":"Article 103256"},"PeriodicalIF":5.8,"publicationDate":"2025-06-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144213028","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Microbiome dynamics and multiscale environmental response patterns of later-diverging coral clade across latitudes, reefs and geomorphological zones in the South China Sea","authors":"Biao Chen ,&nbsp;Lin Liang ,&nbsp;Kefu Yu ,&nbsp;Yuxin Wei ,&nbsp;Xinyue Liang ,&nbsp;Zeming Bao ,&nbsp;Zhiheng Liao ,&nbsp;Xiaopeng Yu ,&nbsp;Zhenjun Qin ,&nbsp;Lijia Xu ,&nbsp;Yongzhi Wang ,&nbsp;Yaru Kang","doi":"10.1016/j.ecoinf.2025.103244","DOIUrl":"10.1016/j.ecoinf.2025.103244","url":null,"abstract":"<div><div>The climatic adaptability and resilience of coral-associated microbiomes are pivotal under the global change. However, the environmental responses and acclimation patterns of microbiome within corals from the latest clades across multiple spatial scales remain unclear. This study analyzed the community and function characteristics of Symbiodiniaceae and bacteria in <em>Lithophyllon scabra</em> (latest-diverging clade of Fungiidae) across latitudes, reefs and geomorphological zones in the South China Sea. The results showed that <em>L.scabra</em> acclimated to environmental variation at multiple spatial scales by establishing specific symbioses with C27 sub-clade. The deterministic assembly of Symbiodiniaceae was associated with nutrient declines at latitudinal scales, while at reefal and geomorphological scales, it is driven by climatic factors and their interactions with local effects, respectively. However, the stochastic process of Symbiodiniaceae was shaped by symbionts dispersal across multiple spatial scales. Notably, environment filtration entirely governed the bacterial assembly process. At latitudinal and reefal scales, the environmental effects and responses pattern of bacterial community aligned with “Pierre Cardin principle” and “Anna Karenina principle”, respectively. Interestingly, bacterial community was enriched with nitrogen-metabolizing taxa and photoautotrophic functions in the lagoon, while exhibiting a higher abundance of heterotrophic functions and antibacterial taxa on the reef slope, which suggests that changes in nutritional patterns and composition of the bacterial community were crucial for the acclimation of <em>L. scabra</em> to distinct geomorphological zones. These results provide novel insights into the environmental interactions and adaptive strategies of the microbiome associated with younger clades of coral across multiple spatial scales in the context of climate change.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"90 ","pages":"Article 103244"},"PeriodicalIF":5.8,"publicationDate":"2025-06-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144213032","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Reliable unmanned aerial vehicle-based thermal infrared target detection method for monitoring Procapra przewalskii in Qinghai","authors":"Guoqing Zhang ,&nbsp;Wei Luo ,&nbsp;Yongxiang Zhao ,&nbsp;Quanqin Shao ,&nbsp;Lin Li ,&nbsp;Keyu Mei ,&nbsp;Guohong Li","doi":"10.1016/j.ecoinf.2025.103209","DOIUrl":"10.1016/j.ecoinf.2025.103209","url":null,"abstract":"<div><div><em>Procapra przewalskii</em> plays a vital role in maintaining ecological balance; however, it faces considerable threats due to habitat degradation and illegal poaching. Monitoring this species using unmanned aerial vehicles (UAVs) has proven to be an effective conservation strategy. A major challenge in UAV-based surveillance of <em>Procapra przewalskii</em> is conducting observations at night or under conditions of poor visible light. To address this issue, this paper presents a thermal infrared (TIR) target monitoring technique using UAVs. This technique employs YOLOv8s as the base model and proposes a multi-frame processing (MFP) method (YOLO-MFP). This method uses the current frame as the primary input and combines optical flow–processed images and background-suppressed images as auxiliary inputs. Background-suppressed images can effectively minimize most background pixels, while regions with high vector values in optical flow–processed images indicate object positions. The model extracts raw feature data, object details, and movement information from these inputs to improve detection performance. Additionally, a small target detection layer is added to reduce missed detections of smaller targets in TIR images while enhancing the overall detection accuracy. Furthermore, the VoVGSCSP module refines the model's neck architecture by effectively merging the feature maps across various stages, reducing computational demands without sacrificing detection precision. Finally, through numerous comparative experiments on our proposed TIR-<em>Procapra przewalskii</em> dataset, YOLO-MFP reaches a mean average precision ([email protected]) value of 96.4 %, precision value of 92.6 %, and recall of 97.0 %, making it superior to the current state-of-the-art models. The importance of this study lies in its enhanced monitoring capabilities for <em>Procapra przewalskii</em>, providing valuable insights for future UAV-based wildlife observation efforts.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"90 ","pages":"Article 103209"},"PeriodicalIF":5.8,"publicationDate":"2025-06-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144213033","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Comparison of climate-phenology-hydrology associations at two long-term studied forest watersheds in subtropical mountainous Taiwan","authors":"Chung-Te Chang ,&nbsp;Jun-Yi Lee ,&nbsp;Jyh-Min Chiang ,&nbsp;Hsueh-Ching Wang ,&nbsp;Cho-ying Huang ,&nbsp;Jr-Chuan Huang ,&nbsp;Chiu-Hsien Wang ,&nbsp;Chun-Wei Tseng","doi":"10.1016/j.ecoinf.2025.103257","DOIUrl":"10.1016/j.ecoinf.2025.103257","url":null,"abstract":"<div><div>Forested watersheds provide clean water and stabilize hydrological services. Associations between vegetation growth and climatic variation significantly influence hydrological regimes, which are region-dependent. However, the climate-phenology-hydrology nexus has rarely been investigated in subtropical forested watersheds, which remain underexplored due to their year-round vegetation activity complicating the phenological shift detection, and highly variable seasonal rainfall introducing uncertainty in hydrological modeling. Understanding these dynamics provides insights into subtropical forests' buffering capacities against climatic fluctuations and their water regulation. This study examined monthly temperature and precipitation impacts on vegetation growth using monthly photosynthetic active vegetation cover fraction (PV) from satellite imagery, assessing the effects of spring and summer rainfall (2001−2020) on vegetation phenology and streamflow in subtropical Fushan (annual precipitation &gt;4000 mm) and tropical Leinhuachi (annual precipitation &gt;2300 mm) Experimental Forests. PV and temperature exhibited linear correlations without time-lag effect (R² = 0.51–0.57, <em>p</em> &lt; 0.001). However, PV and precipitation had no time-lag in Fushan, but demonstrated a log-linear relationship with two-month lag in Leinhuachi (R² = 0.15–0.59, <em>p</em> &lt; 0.001), highlighting rainfall accumulation during the relatively dry season (winter-spring) as critical for vegetation growth. Structural equation modeling (SEM) revealed that an earlier start of growing season (SOS), driven by high spring rainfall (February–March), led to an extended growing season and higher P-Q deficit (precipitation minus runoff) during Leinhuachi's growing season (Goodness of fit index = 0.988, <em>χ</em>² = 0.825, df = 2, <em>p</em> = 0.662). Surprisingly, abundant growing season precipitation had no significant impact on season end, length, or P-Q deficit. These patterns were absent in Fushan (Goodness of fit index = 0.997, <em>χ</em>² = 0.137, df = 2, <em>p</em> = 0.934). Integrating seasonal precipitation variability into watershed management is critical for water security, particularly in drought-prone subtropical regions. Furthermore, incorporating these climate-phenology-hydrology patterns into climate change models will enhance predictions of ecosystem responses, especially where seasonal precipitation affects vegetation productivity, water budgets, and carbon cycling. Understanding subtropical forest regulation of water and carbon cycles is essential to improve climate projections and conservation policies.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"90 ","pages":"Article 103257"},"PeriodicalIF":5.8,"publicationDate":"2025-06-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144212947","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Using correlative science, open access big data and ensemble machine learning to track contamination signals in the wild: A first landscape-scale prediction for the Himalayan vulture (Gyps himalayensis) associated with diclofenac in Asia","authors":"Dikpal Krishna Karmacharya ,&nbsp;Ganesh Puri ,&nbsp;Ganga Ram Regmi ,&nbsp;Madan Krishna Suwal ,&nbsp;Falk Huettmann","doi":"10.1016/j.ecoinf.2025.103243","DOIUrl":"10.1016/j.ecoinf.2025.103243","url":null,"abstract":"<div><div>The Himalayan vulture (<em>Gyps himalayensis</em>) is the largest vulture in central Asia with a wide reach across the tropical mountain parts and landscapes of the Old World. While they co-evolved with humans for millennia, they are now on a decline in most parts of their range, e.g. due to contaminants in the food chain with non-steroidal anti-inflammatory drugs (NSAIDs) like Diclofenac. Summarized with a workflow, here we present the first correlational Big Data mining approach using Open Access Data for vultures and associated GIS layers in the Old World. We used latest machine learning algorithms to obtain the best possible prediction for inference. Due to the established role of Diclofenac as a local extinction factor for vultures we are correlating the best available vulture prediction with the digitally best-available global diclofenac layer. We find that vultures are fully exposed to essentially one of three levels of diclofenac: unknown, lower units and very high amounts. Many remaining vulture presences now correlate with low Diclofenac units whereas high Diclofenac shows little vultures predicted, if at all. We find most of the high risk zones to be located in China (by area and outliers), Mongolia, Pakistan, Afghanistan, Tajikistan and Bangladesh, whereas Nepal for instance seems to be rather low risk. In the absence of mechanistic studies on a larger scale we propose that our pioneering work still represents an underestimate due to several confounding actors not resolved, e.g. farming and high altitude refugia. But it can be used to prioritize, pursue and fine-tune these results, inform conservation and pre-cautionary management, and use our workflow to further study, quantify and safeguard raptors and this species that exemplify such a food chain in the Anthropocene, e.g. through large Diclofenac-free zones.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"90 ","pages":"Article 103243"},"PeriodicalIF":5.8,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144221560","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Leveraging Landsat and Google Earth Engine for long-term chlorophyll-a monitoring: A case study of Lake Balaton's water quality","authors":"Huan Li ,&nbsp;Boglárka Somogyi ,&nbsp;Xiaona Chen ,&nbsp;Zengliang Luo ,&nbsp;Katalin Blix ,&nbsp;Sirui Wu ,&nbsp;Zheng Duan ,&nbsp;Viktor R. Tóth","doi":"10.1016/j.ecoinf.2025.103245","DOIUrl":"10.1016/j.ecoinf.2025.103245","url":null,"abstract":"<div><div>Chlorophyll-a (<em>Chl-a</em>) is one of the critical water quality indicators that shows the eutrophication status of aquatic ecosystems. As the largest lake and a well-known attraction in middle Europe, Lake Balaton contributes 70 % or more of local economy through tourism, while also maintaining a unique biodiversity. Therefore, long-term monitoring of water quality is essential for its effective management. With the longest global environmental record and a preferable spatial resolution, the satellite constellation Landsat is used for retrieving <em>Chl-a</em> in this study. However, the common low-frequent in-situ samplings and ∼ 16-day revisit of Landsat have limited both the quality and applicability of Landsat to <em>Chl-a</em> retrieval. Initially, we trained both linear and several machine learning models using matchups between in-situ measurements and satellite data from Landsat 4–9 missions during 1984 and 2023. To address the imbalanced data problem, which lacks high concentration samples due to the rare blooming events, we extend the time tolerance, incorporate temporal information, which connotes the phenology information, and apply an oversampling technique during the training process. Validated on Lake Balaton, which has a spatiotemporal amplitude of <em>Chl-a</em> concentration ranging from 5 to 260 μg/L since 1980s, Random Forest model has the best accuracy, which shows an R-square 0.86 and RMSE 8.16 μg/L. The oversampling technique improves the accuracy by 9.5 % than the non-oversampled. Leveraging all strategies improves overall accuracy by 21 %. The result also shows a reasonable trade-off via increasing the number of matchups 3 to 8 times by extending the time tolerance from the same day to 3 days regardless of the high variability of <em>Chl-a</em> due to the sinking and floating movement of algae. The enhancement framework can be applied to other lakes, especially for lakes with coarse samplings and wide <em>Chl-a</em> fluctuations. We present an open-source online tool for historical and real-time <em>Chl-a</em> mapping, designed for both experts and the public. With customizable code for global lakes, results are continuously showcased on the HUN-REN Balaton Limnological Research Institute's website and social media.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"90 ","pages":"Article 103245"},"PeriodicalIF":5.8,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144213029","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}