Forest fires are complex spatiotemporal phenomena influenced by local interactions, environmental conditions, and human intervention. In this study, we propose a hybrid modeling framework that integrates a probabilistic cellular automaton (CA) for fire propagation with mobile intelligent agents for active suppression. The fire dynamics are governed by neighbor-dependent stochastic rules modulated by vegetation density and wind intensity, while suppression agents dynamically move toward burning sites and probabilistically extinguish fires. We conduct systematic multi-scenario experiments to examine (1) the impact of wind speed, (2) vegetation density, and (3) the number of suppression agents on fire evolution. Results reveal nonlinear relationships between these parameters and suppression efficiency, including threshold effects and diminishing returns in agent scaling. The model demonstrates how spatial structure, stochasticity, and adaptive control jointly shape fire outcomes. Our framework provides a flexible platform for studying coupled human-environment dynamics and offers insights into optimal resource allocation for wildfire management.
The total number of species on Earth is a fundamental metric in ecology and conservation biology, yet after three centuries of taxonomic effort estimates span two orders of magnitude (2 million to over 1 billion), with little sign of convergence. This persistent uncertainty stems from three intertwined challenges: the preponderance of rare and undersampled species, the methodological incommensurability of extrapolation, decompositional, and molecular approaches, and the lack of a universal species concept for microorganisms. This study proposes the Hierarchical Weighted Cross-Calibration (HWCC) framework, a four-layer probabilistic integration method that synthesizes domain-specific expert estimates, higher-taxon regression constraints, statistical lower bounds (Chao-class estimators), and molecular correction factors derived from eDNA metabarcoding. The seven essential dimensions of the estimation problem are systematically reviewed and quantified: (1) calibration of the known-species baseline, (2) efficacy testing of classical macroecological extrapolations, (3) statistical inference of unseen species, (4) domain-decomposed estimation for major trophic and habitat guilds, (5) quantification of taxonomic dark matter revealed by high-throughput sequencing, (6) dynamic correction for net species loss under contemporary extinction rates, and (7) philosophical and operational reconstruction of the microbial species concept. Monte Carlo propagation of all quantified uncertainties (106 iterations) yields a median global eukaryotic species richness of 71 million, with a 90 % credible interval of 65—78 million. When prokaryotic molecular operational units based on a 95 % average nucleotide identity threshold are included, the interval broadens to 72—86 million. The estimate represents a 7- to 8-fold upward revision from the widely cited 8.7 million and implies that the denominator for current extinction-rate calculations, and thus the magnitude of unrecognized "dark extinctions", has been underestimated by an order of magnitude. The HWCC framework is openly structured for iterative refinement with new data and provides a probability density distribution rather than a single-point estimate, enabling risk-explicit incorporation into global conservation targets and biodiversity scenarios.
The geographical patterning of species diversity, most famously the latitudinal gradient, one of the oldest and most celebrated patterns in ecology, still lacks a unified mechanistic explanation. Existing hypotheses, including the species-energy hypothesis, the environmental heterogeneity hypothesis, the area hypothesis, neutral theory, metabolic theory, maximum entropy theory, and historical-evolutionary hypotheses, each capture a critical dimension of diversity generation. Yet none alone can explain why regions with identical energy inputs can differ so dramatically in species richness. After systematically reviewing these classical and cutting-edge theories, I propose a novel integrative framework: the Niche-Energy-Time triadic synergy hypothesis (NET hypothesis) in present paper. The NET hypothesis starts from three irreducible ultimate constraints: a thermodynamic constraint, available energy flux (E) sets the upper limit on the total biomass and number of individuals a community can sustain; a structural constraint, multidimensional niche space volume (H) determines the fineness with which that energy flow can be partitioned among species; and a historical constraint, the effective evolutionary and community assembly time (T) determines the degree to which that niche space has been filled. I argue that species diversity is an emergent outcome of these three constraints acting as a serial filter not a linear function of any single factor. I present the core mathematical structure of the NET hypothesis, demonstrate its logical necessity by deriving it from population energy allocation, the niche-width-species-number trade-off, and the macro-dynamics of speciation-extinction balance, and show its power to unify a wide range of classical diversity patterns, including the latitudinal gradient, elevational patterns, and island species-area relationships, as well as anomalous cases. The NET hypothesis does not overturn existing hypotheses but embeds energy, heterogeneity, area, and time into the E, H, and T dimensions, revealing the synergistic mechanisms by which they act as necessary but not sufficient conditions. It provides a testable, quantifiable, mechanistic foundation for predicting biodiversity change and guiding conservation planning.
A hierarchical indicator system and scoring sytem to assess plant species' adaptation to climate change was constructed in present study. The AI-driven assessment tool based on the systems was developed. It features with (1) Multi-dimensional adaptive capacity analysis. (2) Support for multiple AI providers (DeepSeek, Google Gemini). (3) Real-time progress tracking. (4) Comprehensive scoring system. (5) Actionable recommendations for conservation and management. By integrating physiological, biological, morphological, and genetic perspectives, it offers actionable insights for conservation planning and ecosystem management.
Understanding the global distribution of species is crucial for their conservation, utilization, and management of the species. In present study, the AI-driven web tool, Species Distribution Finder, was developed. The Species Distribution Finder is a front-end HTML/JavaScript application that enables users to query an AI model (DeepSeek or Google Gemini) to obtain detailed species geographic distribution data. The Species Distribution Finder guides the AI to return structured, exhaustive lists of species occurrence locations (countries, provinces/states, counties, cities, etc.). It acts as a structured front-end AI querying interface which uses precise prompt engineering for scientific comprehensiveness, provides fluid, user- friendly interaction, and enables reproducible, AI-driven biodiversity data exploration. It can be used as a real-time finder for global geographic distribution of any species.
Deforestation, driven by rapid urbanization, industrialization, and other developmental activities, poses a significant threat to wildlife by disrupting natural habitats and forcing various wildlife species to migrate in search of more suitable and safer environments. This uncontrolled and often irregular migration frequently results in serious ecological imbalances, rising instances of human-wildlife conflicts, and the potential endangerment or extinction of vulnerable species. To gain a deeper understanding of the complex relationship between deforestation and emerging wildlife migration patterns, this paper presents a detailed nonlinear mathematical model that carefully examines the consequences of habitat degradation on the movement of wildlife across increasingly fragmented ecosystems. The model includes three key and interdependent variables: the cumulative density of forestry resources, the extent and intensity of developmental activity, and the population density of impacted wildlife species. It is constructed using a set of nonlinear ordinary differential equations and analyzed through stability theory to determine both equilibrium conditions and long-term dynamics. The results clearly demonstrate that as deforestation increases due to intensified developmental activities, both forest biomass and wildlife populations undergo a marked decline, resulting in elevated migration rates. Numerical simulations support these conclusions, underlining the critical need for sustainable development practices to reduce habitat destruction and preserve ecological balance.
Climate change poses unprecedented threats to global biodiversity, necessitating advanced computational frameworks for predicting species distributions and ecosystem responses under future climate scenarios. Traditional species distribution models (SDMs) and mechanistic approaches lack the capacity to capture complex nonlinear ecological dynamics while remaining interpretable to conservation practitioners (Christin et al., 2019). This paper presents a novel hybrid framework integrating Physics-Informed Neural Networks (PINNs) with Graph Neural Networks (GNNs), enhanced by multi-scale attention mechanisms and Bayesian uncertainty quantification (Wesselkamp et al., 2024). Our approach embeds mechanistic ecological constraints directly into neural architectures while explicitly modeling species interaction networks as graph-structured data (Anakok et al., 2025). We evaluate the hybrid PINN-GNN model against traditional SDMs, standard deep neural networks, and standalone PINN/GNN approaches using a multi-regional dataset spanning 225 species across diverse ecosystems. Results demonstrate superior predictive performance: 92% accuracy (vs. 81% for standard DNNs and 72% for traditional SDMs), RMSE of 0.08 (71% improvement over traditional methods), and AUC-ROC of 0.95. Explainable AI analysis via SHAP values (He et al., 2022) identifies temperature (0.42), habitat fragmentation (0.35), and precipitation (0.28) as the most influential environmental drivers. Climate change projections under SSP2-4.5 and SSP5-8.5 scenarios predict range shifts of 82-195 km by 2100, with 73% of species experiencing net range contractions. Bayesian uncertainty quantification reveals growing epistemic uncertainty (0.03-0.08) as ecosystems enter novel climates (Olivier et al., 2021). This research advances computational ecology by providing an interpretable, mechanistically-grounded, uncertainty-aware framework suitable for biodiversity conservation planning in the Anthropocene.
Acid rain formation in the atmosphere is a serious environmental hazard that causes enormous ecological harm. It is formed by the interaction of air pollutants with vapour clouds. It threatens biodiversity, damages ecosystems, and has a negative effect on plant growth. In this paper, we propose a nonlinear mathematical model to examine how acid rain affects the growth of plant biomass in a habitat. The model takes into account five important factors: the density of plant biomass, the density of human population, the cumulative density of sources that release pollutants, the cumulative concentration of pollutants, and the cumulative concentration of acid rain. The model makes the assumptions of logistic growth for both human population density and plant biomass, which is adversely affected by acid rain. The model analysis shows that increasing the concentration of acid rain leads in a considerable drop in plant biomass density. The model analysis, which employs the stability theory of differential equations, shows that a substantial decrease in plant biomass density at equilibrium results from rising acid rain concentrations or human population density. These findings are supported by numerical simulations, which show that acid rain has a considerable effect on plant biomass growth. This study provides a framework for understanding the ecological consequences of acid rain.
This study aims to integrate open-source computational tools with ecological biomonitoring by harnessing the capabilities of RStudio and the DiaThor package. The goal is to simplify the calculation of diatom-based indices and generate insightful visualizations for water quality assessment. Using RStudio as the platform and the DiaThor package as the computational engine, a case study was conducted on water quality datasets. Various ecological indices were computed including IPS, TDI, SLA, and others. The manuscript provides an end-to-end demonstration of data formatting, function usage, and graphical outputs for visual interpretation. The study successfully generated multiple plots such as bar plots, heatmaps, and radar diagrams, each representing the variation in diatom-based indices across multiple sample sites. The visualization outputs made it easier to detect ecological gradients and interpret water quality conditions efficiently. DiaThor in RStudio offers an efficient and reproducible method for ecological data processing. Its integration with R's visualization ecosystem enhances data clarity and enables automation in environmental monitoring workflows. This paper underscores the potential of open-source packages to revolutionize diatom-based assessment practices.
In present study a system of hierarchical indicators to assess animal species' adaptation to climate change was constructed, and the AI-driven assessment tool was developed. It is a single-page web tool that: (1) Collects species and context information from the user. (2) Lets the user choose an AI provider (DeepSeek or Google Gemini) and enter an API key. (3) Builds a detailed prompt requesting a JSON-formatted climate adaptation assessment. (4) Sends the prompt to the selected provider's REST API. (5) Parses and validates the JSON-like response. (6) Renders the results into a rich UI, grouped into: Physiological indicators, Morphological indicators, Behavioral indicators, Genetic indicators, Summary assessment and conservation recommendations. With this AI tool, we can easily judge the real-time vulnerability of the given animal species to climate change and obtain conservation recommendations.
Aquatic ecosystems are highly sensitive to changes in environmental conditions, making it essential to identify the key factors that influence the dynamics of species populations. This study introduced a nonlinear mathematical model, analyzed it, and identified key sensitive parameters that were used in assessing the impact of water pollution on aquatic ecosystems. Global sensitivity analysis was conducted to determine the parameters significantly impacting aquatic species populations. Parameters were estimated using the least squares method, while sensitivity analysis was performed via Partial Rank Correlation Coefficient (PRCC) and Latin Hypercube Sampling (LHS). Parameters related to organic pollutant growth rates, pollutant absorption rates, and oxygen penetration were identified as positively affecting aquatic species populations by enhancing nutrient availability and metabolic activity. Conversely, competition and inorganic pollutant discharge were found to impact aquatic populations negatively. These findings highlight the critical role of managing sensitive parameters such as pollutants and competitive interactions to maintain and improve the health of aquatic ecosystems.
Dissolved Oxygen (DO) serves as a crucial measure of water quality, imperative for both aquatic life and human consumption. The application of deep learning, particularly through data-driven predictions, offers a robust tool for estimating DO concentrations. Enhanced precision is achieved by fine-tuning hyperparameters. Bayesian optimization methods are amongst those noteworthy for their effectiveness. This study focuses on predicting DO levels using a Deep Neural Network model. The study uses Bayesian optimization to refine hyperparameters for the best model setup, comparing the results with a baseline model using default settings. Results indicate that the Bayesian-optimized model outperforms the baseline. The findings underscore the pivotal role of Bayesian optimization in elevating model performance, exhibiting robust generalization capabilities while significantly reducing the need for manual parameter tuning. This successful application underscores a substantial methodological advancement in environmental management, particularly in predictive modelling for indicators of aquatic ecosystem health.
Mangrove ecosystem plays a crucial role in promoting biodiversity and mitigating climate change impacts. Calatagan, Batangas as a corridor of the Verde Island Passage, Philippines - the center of the center of marine shore fish biodiversity in the world, has a huge potential for carbon storage because of its rich biodiversity. The biodiversity of mangroves and carbon sequestration estimation are important aspects in conservation planning to reduce the impact of climate change. This research assessed the carbon storage of a community-managed mangrove forest in terms of vegetation and soil carbon. The transect plot technique was employed to assess the forest structure and carbon stock. Vegetative carbon was estimated in terms of the aboveground and belowground biomass of the mangroves using allometric models. Physico-chemical analyses of the sediment were conducted to calculate the soil organic carbon density. Based on the findings, Calatagan Mangrove Forest Conservation Park has 12 species. In terms of edaphic patterns, soil organic carbon showed inverse relationship with soil bulk density. Among the species, Avicennia marina and Sonneratia alba had high amounts of vegetative carbon stock. Furthermore, it was found that higher carbon stock is accumulated in the soil than vegetative carbon. From these results, the blue carbon approach to community-based conservation and management of mangrove forests is endorsed. This implies that more locally managed marine protected areas (MPA) should be established to promote and strengthen the blue carbon approach as a nature-based solution to climate change. It also suggests that carbon sequestration must be an integral component of the community-based mangrove conservation and restoration plans towards climate change mitigation and adaptation.
This article presents the software and hardware components of an open-source system designed for water temperature monitoring. The system can handle measurements from up to eight sensors at user-defined time intervals. It was designed with the waterproof temperature sensor model DS18B20, an Arduino board and a computer. The Arduino-based program reads the sensors data and transmits the temperature values to the computer in real-time via a serial communication. The computer software has an intuitive user interface to ensure an easy operation, and it displays the temperatures numerically, generates graphs, and allows the user to save the data to a disk at any time. Besides the Arduino and the sensors, the hardware setup requires only a single additional resistor making assembly straightforward. This article details the Arduino and the computer source code programs, and the satisfactory system results have motivated the work presentation.
Integrated Pest Management (IPM) is an ecosystem-based strategy that focuses on sustainable control of pests through a combination of techniques. In present study, an indicator system, which is a hierarchical system, for eco-sustainability assessment of Integrated Pest Management techniques is proposed. The indicator system is based on various IPM techniques in which an indicator represents a category of IPM techniques (external interventions). The eco-sustainability assessment follows such criteria as the impact of external interventions on the ecosystem and environment (e.g., ecosystem completeness, environmental impact, and human health impact, etc.), the intensity and frequency of external interventions, and the sustainability of IPM. In the assessment system, all categories of IPM techniques are scored and the weighted score for IPM techniques used is calculated for the assessment of IPM sustainability. A calculator is developed for assessment. The calculator is web browser based that includes both online and offline versions and can be used on web browsers. The system can be used to assess an IPM programme and compare between IPM programmes, or used as a tool for IPM teaching and training.
Geographic Information Systems (GIS) have dramatically altered the landscape of spatial data analysis, which have allowed practitioners and scholars to gain important insights in a variety of domains, including emergency response, environmental management, and urban planning. Central to the functionality of GIS are two integral components: maps and models. Although maps have historically been used to graphically and statically depict spatial data, GIS models go beyond static representations to turn maps into dynamic, predictive tools. These models help predict and analyze spatial dynamics across time by simulating real-world phenomena. These models allow for simulations of real-world phenomena, aiding in forecasting and analyzing spatial processes over time. This paper explores the evolution of GIS, focusing on the shift from traditional cartographic maps to dynamic models, highlighting how these advancements have revolutionized spatial analysis and influenced real-world applications. An understanding of how GIS is influencing the future of spatial data interpretation is provided by the exploration of the interaction between maps and models.
In present study, a calculator, probDistriCal, was developed for probability distributions. In the calculator, the probability distributions as normal distribution, t distribution, F distribution, χ2 distribution, exponential distribution, power law distribution, Weibull distribution, Zhang distribution, binomial distribution, Poission distribution, and negative binomial distribution, etc., were available for use. Given a random value, the corresponding probability for a known probability distribution can be calculated. The calculator is web browser based that includes both online and offline versions and can be used on various computing devices (PCs, iPads, smartphones, etc.), operating systems (Windows, Mac, Android, Harmony, etc.) and web browsers (Chrome, Firefox, etc).
The global change of the planet's climate is associated with increased concentration of greenhouse gases in the atmosphere, which is the result of irrational human economic activity. Wetlands are a natural and efficient store of carbon dioxide. It covers only about 6% of the land surface. Peatlands contain one-third of all soil carbon, or 600 billion tons, which is two times more than the entire global forest biomass pool. Only ocean sediments contain more carbon. Peatlands in the boreal zone, characterized by snowy winters and short warm summers, contain on average seven times more carbon per hectare than any other ecosystem, and ten times more in the tropics. Restoration of drained wetlands to enable their use for carbon farming is also advisable, because it reduces carbon dioxide emission caused by microbial oxidation of peat and by wildfires in the drained areas. Therefore, comprehensive research on the role of drained wetlands in capturing and storing greenhouse gases is crucial for the sustainable management of these valuable ecosystems. Identifying the most promising areas for carbon farming is an important challenge for both global and regional studies. In 2022, the Basyanovsky and Koksharovsko-Kombaevsky drained peatlands located in the Sverdlovsk region (the Middle Urals) were studied as a part of reconnaissance work aimed at finding and selecting peatlands suitable for carbon farming. These peatlands are currently not used for peat harvesting and undergo active natural restoration. The area of secondary rewetting within the Koksharovsky peatland was also studied. On the territory of the Koksharovo-Kombayevsky peatland, the species composition of vegetation in both woody and grass-shrub strata, the composition of peat deposits, and the rate of carbon accumulation were evaluated. Based on the results of this research, the most promising peatland for carbon farming and further study of greenhouse gas emissions appeared to be the Koksharovsko-Kombayevsky peatland.
Sexually transmitted infections (STIs) are diseases transmitted mostly through unprotected sex with an infected partner. STIs can be transmitted to an infant before or during childbirth. More than one million sexually transmitted infections (STIs) are acquired every day worldwide. In the most recent years, the prevalence of STIs reached approximately 20% among Tanzanian older adults living in metropolitan areas. If not treated properly and on time, STIs can have severe consequences, including infertility, sterility, increased susceptibility to more serious diseases such as the Human Immunodeficiency Virus (HIV), and even death. However, stigma and shame associated with STIs remain significant barriers to proper diagnosis and timely treatment, leading many patients to face increased risks. The purpose of this paper is to present a machine-learning model for early detection of sexually transmitted infections that was developed. The developed model can be deployed into health systems for self-diagnosis to remove communication barriers between sexual health clinics and STI patients. The study used a quantitative research method and got its dataset of 13,335 records from the Government of Tanzania Health Operations Management Information System (GoT-HoMIS) in areas with many STI cases. This was done by using surveys and questionnaires to get the data. The dataset was split into a 70%:15%:15% ratio for training, testing, and validation, respectively, and five machine learning algorithms were evaluated: AdaBoost, Support Vector Machine, Random Forest, Decision Tree, and Stochastic Gradient Descent. Based on evaluation metrics, the AdaBoost model was identified as the best-performing model, achieving an accuracy of 97.45%, an F1 score of 97.7%, and the Receiver Operating Characteristics Area Under the Curve (ROC-AUC) with a higher true positive rate and a lower false positive rate. The study recommends integrating a machine learning model into healthcare systems to detect STIs early, improve medical care, reduce disease progression, and remove stigmatisation barriers. Also, it can provide insights into infection patterns, allowing practitioners to adapt their responses. Machine learning-based solutions in mobile apps and telemedicine systems promote early testing and treatment.
This study presents a design of a Data-Driven software application for identification, population monitoring, and risk assessment for lions in Serengeti Tanzania. Lions' populations have been declining due to poaching, overhunting, and other ecosystem factors resulting in unmet demands for tourism and ecological balance. Data-driven techniques can lower the negative consequences by providing mechanisms for lions' management, risk assessment, and monitoring in selected wildlife reserves. Lion's whisker spots, poaching rates, prey availability, human-conflict incidences, and pride size are key elements for achieving management, identification, monitoring, and risk assessment for lions. The software application design aimed at providing conceptual and logical requirements for the development of the application that will enhance lions' monitoring and management efforts to protect their existence and contribution to the ecosystem. The study was conducted in the Serengeti ecosystem, including ecologists from the Tanzania Wildlife Research Institute Serengeti Wildlife Research Center, and information systems analysts. Through a mixed research methods approach, qualitative methods and incremental prototyping software development life cycle model were used to develop the specific requirements. Unified Modeling Language (UML) was used to model the requirements and led to the realization of design diagrams: application framework, database design, and artificial intelligence model workflows. The application should equip ecologists with tools to add and identify specific lions, monitor sightings, estimate population trends, assess risks for individual lions, and produce reports on monitoring and sightings. This design serves as a foundation for developing the data-driven software application for identification, population monitoring, and risk assessment for lions in Serengeti National Park Tanzania which will enhance monitoring and management activities of lions' population non-invasively.
