Distinct Modeling Approaches Research Articles

An exploration of RSM, ANN, and ANFIS models for methylene blue dye adsorption using Oryza sativa straw biomass: a comparative approach

This study focused on simulating the adsorption-based separation of Methylene Blue (MB) dye utilising Oryza sativa straw biomass (OSSB). Three distinct modelling approaches were employed: artificial neural networks (ANN), adaptive neuro-fuzzy inference systems (ANFIS), and response surface methodology (RSM). To evaluate the adsorbent’s potential, assessments were conducted using Fourier-transform infrared spectroscopy (FTIR) and scanning electron microscopy (SEM). The evaluation of RSM, ANN, and ANFIS included the quantification of R2, mean squared error (MSE), root mean square error (RMSE), and mean absolute error (MAE) metrics. The regression coefficients from the process modelling demonstrated that RSM (R2 = 0.9216), ANN (R2 = 0.8864), and ANFIS (R2 = 0.9589) all accurately predicted MB adsorptive removal. However, comparative statistical analysis revealed that the ANFIS model exhibited superior accuracy in data-based predictions compared to ANN and RSM models. The ideal pH for MB adsorption utilizing OSSB was established as 7. Additionally, favourable outcomes were obtained with 60-minute contact durations, 20 mg adsorbent quantities, and temperatures of 30 °C. The pseudo 2nd -order kinetic model for MB adsorption by OSSB was confirmed. The equilibrium data exhibited a superior fit with the Langmuir isotherm model in comparison to the Freundlich model. The thermodynamic adsorption parameters, including (∆G = -9.1489 kJ/mol), enthalpy change (∆H = -1457.2 kJ/mol), and entropy change (∆S = -19.03 J mol−1 K−1) indicated that the adsorption of MB onto the OSSB surface is exothermic and spontaneous under the experimental conditions. This research effectively showcased the potential of RSM, ANN, and ANFIS in simulating dye removal using OSSB. The generated parameter data proved valuable for the design and control of the adsorption process.

From situated knowledges to situated modelling: a relational framework for simulation modelling

ABSTRACT In this paper we extend the use of a relational approach to simulation modelling, a widely used knowledge practice in sustainability science. Among modellers, there is awareness that model results can only be interpreted in view of the assumptions that inform model construction and analysis, but less systematic questioning of those assumptions. Moreover, current methodological discussions tend to focus on integrating social and ecological dynamics or diverse knowledges and data within a model. Yet choices regarding types of modelling, model structure, data handling, interpretation of results and model validation are not purely epistemic. They are entangled with values, contexts of production and use, power relations, and pragmatic considerations. Situated Modelling extends a relational understanding of the world to scientific knowledge production and with that to modelling itself in order to enable a systematic interrogation of these choices and to research social-ecological transformations relationally. To make tangible the situatedness of simulation modelling, we build on existing practices and describe the situatedness of three distinct modelling approaches. We then suggest four guiding principles for Situated Modelling: 1. attending to the apparatus of knowledge production that is socially and materially embedded and produced by e.g. research infrastructures, power relations, and ways of thinking; 2. considering how agency is distributed between model, world, data, modeller in model construction; 3. creating heterogenous collectives which together occupy the formerly individualised subject position; and 4. using agonism as an epistemic virtue to retain and work with significant differentiations of social-ecological dynamics throughout the modelling process.

Analysis and Prediction of Wear in Interchangeable Milling Insert Tools Using Artificial Intelligence Techniques

Milling machines remain relevant in modern manufacturing, with tool optimization being crucial for cost reduction. Inserts for compound cutting tools can reduce the cost of operations by optimizing their lifespan. This study analyzes the flank wear of cutting tools in milling machines, with an emphasis on evaluating different approaches to predict their lifespan. It compares three distinct modeling approaches for predicting tool lifespan using algorithms: traditional ensemble methods (Random Forest, Gradient Boosting) and a deep learning-based LSTM network. Each model is evaluated independently, and this comparative analysis aims to determine which modeling strategy best captures the intricate interactions between various process variables affecting tool wear. This method ensures greater efficiency and accuracy than conventional techniques, providing a scalable, resource-efficient solution for reliable and insightful tool wear predictions. The results obtained from the dataset of an insert tool can be extrapolated to other milling inserts and demonstrate the progression of tool wear over time under varying cutting parameters, providing critical insights for optimizing milling operations. The integration of uncertainty awareness in the predictive outputs is a unique feature of this research and enhances decision-making for smarter manufacturing. This proactive approach enhances operational efficiency and reduces overall production costs. Furthermore, the data-driven, AI-centric methodology developed in this study offers a transferable approach that can be adapted to other machining processes, advancing state-of-the-art tool wear prediction.

Mapping illegal trade routes of live cheetahs from the Horn of Africa to the Arabian Peninsula.

Less than 7000 cheetahs (Acinonyx jubatus) persist in Africa. Although human-wildlife conflict, habitat degradation, and loss of prey are major threats to cheetah populations, illegal trade in live cubs for pets may have the most significant impact on populations in the Horn of Africa. We developed a novel, stepwise decision support tool to predict probable trafficking routes by leveraging the power of distinct modeling approaches. First, we created a cheetah habitat suitability index (HSI) to determine where source cheetah populations may occur. We then created a trafficking network model linking known and predicted cheetah populations with documented destinations in the Arabian Peninsula. A significant area in Eastern Ethiopia and Northern Somalia was estimated to harbor undocumented cheetahs. When these predicted populations were used as a supply source, the trafficking network model showed multiple routes passing through Somaliland and across the Gulf of Aden to Yemen, supporting the notion that undocumented cheetahs may be supplying pet market demands. Though we demonstrate how our decision support tool can inform law enforcement, conservation strategies, and community engagement, we caution that our results are not fully validated due to limited accessibility, alternative trafficking routes, and the cryptic nature of illegal wildlife trade.

Modeling Ordinal First Return Dynamics from Time Series

We examine the family of ordinal First Return Maps (FRMs) of a dynamical system time series to model the system’s dynamics. We consider two distinct modeling approaches using the same modeling method and compare their outcomes to ascertain the prerequisites for each to function and which approach is more effective in different circumstances. The first approach involves short-term modeling of the entire flow, followed by identifying the ordinal sections on the predicted flow. The second approach relies on locating the ordinal sections within the reference time series and using the data from each section independently to build a model that can produce a long-term prediction for each section’s FRM. We demonstrate the effectiveness of these approaches on two well-known dynamical systems, the Lorenz and Rössler equations, emulating data deficiency by selecting a low sampling rate. We consider the effect of downsampling on the shape of the FRMs to choose suitable ordinal sequences to use in creating Poincaré sections from the time series. Following that, we propose a dissimilarity measure between the reference and predicted FRMs to assess the model quality. We then compare the reference and predicted FRMs using the introduced dissimilarity criterion to determine which modeling approach performs better in each situation. Ultimately, we classify the FRMs situation into continuous and piecewise continuous scenarios. Based on our observations from the Lorenz and Rössler dynamical systems, we found that when the system’s FRMs are piecewise continuous, it is more effective to model the entire flow. However, when the FRMs are continuous, our choice of modeling approach depends on the criterion that the predicted FRM is more similar to the reference.

Semi-analytical solutions for land subsidence due to groundwater withdrawal

This paper introduces novel semi-analytical models tailored for estimating land subsidence resulting from groundwater extraction in confined aquifers. These models offer high scalability, allowing them to be applied to various well configurations and pumping schedules. Their development involves the numerical integration of two key analytical solutions: the “nucleus of strain” (NoS) (Mindlin and Chen, 1950), which represents a localised zone within the aquifer where a unit change in pore pressure leads to deformation and subsequent surface displacement, and the classic Theis equation (Theis, 1935) for the pore pressure changes induced by a constant-rate well pumping from a laterally unbounded aquifer. These integrations yield surface displacement components, both horizontal and vertical, expressed as functions of two dimensionless spatial–temporal variables, which encompass aquifer depth, thickness, well placement, pumping schedules, and critical hydro-geomechanical parameters like hydraulic conductivity, porosity, vertical compressibility, and water compressibility. Proposed are two distinct modelling approaches: one employing a lookup table (LT) derived from numerical integration results, and the other providing direct closed-form surface displacement solutions by fitting LT data with “hinge models”, which use piecewise-linear functions linked by sigmoidal curves for computational efficiency. In both cases, surface displacement components are estimated by plugging in the dimensionless variables. Conditions of variable pumping from multiple wells can be addressed by applying superposition of solutions. In essence, these semi-analytical models offer swift computational capabilities for understanding and forecasting land subsidence dynamics. Their scalability makes them adaptable to a wide array of well configurations and scheduling scenarios, rendering them valuable for numerous applications. They are particularly significant for providing preliminary estimates of the impacts of groundwater development, conducting “what-if” tests, and performing sensitivity analyses to identify key factors affecting land subsidence risk. This underscores the importance of these models in sustainable groundwater resource management and in mitigating land subsidence and its associated consequences.

A comprehensive review of deep learning in EEG-based emotion recognition: classifications, trends, and practical implications

Emotion recognition utilizing EEG signals has emerged as a pivotal component of human–computer interaction. In recent years, with the relentless advancement of deep learning techniques, using deep learning for analyzing EEG signals has assumed a prominent role in emotion recognition. Applying deep learning in the context of EEG-based emotion recognition carries profound practical implications. Although many model approaches and some review articles have scrutinized this domain, they have yet to undergo a comprehensive and precise classification and summarization process. The existing classifications are somewhat coarse, with insufficient attention given to the potential applications within this domain. Therefore, this article systematically classifies recent developments in EEG-based emotion recognition, providing researchers with a lucid understanding of this field’s various trajectories and methodologies. Additionally, it elucidates why distinct directions necessitate distinct modeling approaches. In conclusion, this article synthesizes and dissects the practical significance of EEG signals in emotion recognition, emphasizing its promising avenues for future application.

Cybercrime Intention Recognition: A Systematic Literature Review

Cybercrime Intention Recognition: A Systematic Literature Review

Assessing the reliability of a physical-based model and a convolutional neural network in an ungauged watershed for daily streamflow calculation: a case study in southern Portugal

The main goal of this study was to estimate inflows to the Maranhão reservoir, southern Portugal, using two distinct modeling approaches: a one-dimensional convolutional neural network (1D-CNN) model and a physically based model. The 1D-CNN was previously trained, validated, and tested in a sub-basin of the study area where observed streamflow values were available. The trained model was here subject to an improvement and applied to the entire watershed by replacing the forcing variables (accumulated and delayed precipitation) to make them correspond to the values of the entire watershed. The same way, the physically based MOHID-Land model was calibrated and validated for the same sub-basin, and the calibrated parameters were then applied to the entire watershed. Inflow values estimated by both models were validated considering a mass balance at the reservoir. The 1D-CNN model demonstrated a better performance in simulating daily values, peak flows, and the wet period. The MOHID-Land model showed a better performance in estimating streamflow values during dry periods and for a monthly analysis. Hence, results show the adequateness of both modeling solutions for integrating a decision support system aimed at supporting decision-makers in the management of water availability in an area subjected to increasing scarcity.

Toward Predicting Interfacial Tension of Impure and Pure CO2-Brine Systems Using Robust Correlative Approaches.

In the context of global climate change, significant attention is being directed toward renewable energy and the pivotal role of carbon capture and storage (CCS) technologies. These innovations involve secure CO2 storage in deep saline aquifers through structural and capillary processes, with the interfacial tension (IFT) of the CO2-brine system influencing the storage capacity of formations. In this study, an extensive data set of 2811 experimental data points was compiled to model the IFT of impure and pure CO2-brine systems. Three white-box machine learning (ML) methods, namely, genetic programming (GP), gene expression programming (GEP), and group method of data handling (GMDH) were employed to establish accurate mathematical correlations. Notably, the study utilized two distinct modeling approaches: one focused on impurity compositions and the other incorporating a pseudocritical temperature variable (Tcm) offering a versatile predictive tool suitable for various gas mixtures. Among the correlation methods explored, GMDH, employing five inputs, exhibited exceptional accuracy and reliability across all metrics. Its mean absolute percentage error (MAPE) values for testing, training, and complete data sets stood at 7.63, 7.31, and 7.38%, respectively. In the case of six-input models, the GEP correlation displayed the highest precision, with MAPE values of 9.30, 8.06, and 8.31% for the testing, training, and total data sets, respectively. The sensitivity and trend analyses revealed that pressure exerted the most significant impact on the IFT of CO2-brine, showcasing an adverse effect. Moreover, an impurity possessing a critical temperature below that of CO2 resulted in an elevated IFT. Consequently, this relationship leads to higher impurity concentrations aligning with lower Tcm values and subsequently elevated IFT. Also, monovalent and divalent cation molalities exhibited a growing influence on the IFT, with divalent cations exerting approximately double the influence of monovalent cations. Finally, the Leverage approach confirmed both the reliability of the experimental data and the robust statistical validity of the best correlations established in this study.

Analysis of loss of flow without scram test in the FFTF reactor – Part I: Preparation of neutronics data

This study presents a benchmark analysis of an unprotected loss of flow transient in a sodium-cooled fast reactor at the Fast Flux Test Facility (FFTF), carried out as part of an IAEA coordinated research project. Three codes, namely Serpent (Monte Carlo), DYN3D (3D nodal diffusion) and ATHLET (system thermal hydraulics), were employed in the benchmark exercise. Two distinct modeling approaches were utilized: 1) stand-alone ATHLET with point kinetics (PK) and system thermal hydraulics (TH); and 2) coupled DYN3D/ATHLET with spatial kinetics (SK) and system TH. Neutronics data essential for both approaches were generated using Serpent. The study is organized into three parts.Part I presents a summary of the preparation of neutronics data for PK or coupled SK/TH simulations and includes the outcomes of the static neutronics stage of the benchmark. The main focus lies on verifying the cross-section generation method for DYN3D by comparing its results against the reference Monte Carlo solutions obtained with Serpent.Part II will provide a detailed description of the ATHLET TH model of the system. This model will be thoroughly evaluated by comparing the results with the ANL benchmark solution and experimental data for the transient. Furthermore, a sensitivity analysis will be conducted to explore various modeling options and assess their impact on the simulation results.Part III will showcase the transient calculation results using the two modeling approaches. Additionally, an adaptive decay heat model for nodal codes will be introduced. The performance of both modeling approaches will be assessed by comparing their results to the available experimental data.

Sustained beneficial infections: priority effects, competition, and specialization drive patterns of association in intracellular mutualisms

Intracellular symbioses provide a useful system for exploring evolutionary and ecological forces that shape mutualistic partnerships. Within- and among-host competitiveness driven by different strategies that symbionts adopt as they transfer materials to the host across a sub-cellular membrane might explain patterns of host:symbiont association observed in natural systems. We tested the hypothesis that different translocation strategies employed by symbionts affect their ability to occupy host habitats using two distinct modeling approaches. The first involved constructing a deterministic, Lotka-Volterra-type model with two symbiont species competing for access to a single host. The model recovered expected behaviors of co-occupancy/coinfection as well as competitive exclusion. However, a specialization coefficient allowed advantages to accrue to one of the symbionts and permitted otherwise inferior competitors to displace superior competitors. The second approach involved developing and implementing a detailed, highly configurable, and realstic agent-based model (ABM), facilitating experimentation of multiple symbiont strategies in competition simultaneously. The ABM emphasizes bidirectional movement of materials between symbiont and host (e.g., photosynthate from algae to heterotrophic host). Competitive interactions between symbionts based on simple strategies led to exclusion of the inferior symbiont or co-occupancy of the host. As in the first model, inferior competitors could overtake superior competitors when “affinity” terms (i.e., specialization) were included in the model. Both models lay bare the importance of coevolutionary specialization as a selectively advantageous strategy, and they offer a new conceptual framework for interpreting the dynamic patterns observed in extant host and mutualist associations by challenging the idea of “host control” of outcomes, and identifying specific points where coevolutionary specialization might accrue.

In Silico Clinical Trials: Is It Possible?

Modeling and simulation (M&S), including in silico (clinical)trials, helps accelerate drug research and development and reduce costs and have coined the term "model-informed drug development (MIDD)." Data-driven, inferential approaches are now becoming increasingly complemented by emerging complex physiologically and knowledge-based disease (and drug) models, but differ in setup, bottlenecks, data requirements, and applications (also reminiscent of the different scientific communities they arose from). At the same time, and within the MIDD landscape, regulators and drug developers start to embrace in silico trials as a potential tool to refine, reduce, and ultimately replace clinical trials. Effectively, silos between the historically distinct modeling approaches start to break down. Widespread adoption of in silico trials still needs more collaboration between different stakeholders and established precedence use cases in key applications, which is currently impeded by a shattered collection of tools and practices. In order to address these key challenges, efforts to establish best practice workflows need to be undertaken and new collaborative M&S tools devised, and an attempt to provide a coherent set of solutions is provided in this chapter. First, a dedicated workflow for in silico clinical trial (development) life cycle is provided, which takes up general ideas from the systems biology and quantitative systems pharmacology space and which implements specific steps toward regulatory qualification. Then, key characteristics of an in silico trial software platform implementation are given on the example of jinkō.ai (nova's end-to-end in silico clinical trial platform). Considering these enabling scientific and technological advances, future applications of in silico trials to refine, reduce, and replace clinical research are indicated, ranging from synthetic control strategies and digital twins, which overall shows promise to begin a new era of more efficient drug development.

Machine learning-based investigation of the relationship between immune status and left ventricular hypertrophy in patients with end-stage kidney disease.

Left ventricular hypertrophy (LVH) is the most frequent cardiac complication among end-stage kidney disease (ESKD) patients, which has been identified as predictive of adverse outcomes. Emerging evidence has suggested that immune system is implicated in the development of cardiac hypertrophy in multiple diseases. We applied machine learning models to exploring the relation between immune status and LVH in ESKD patients. A cohort of 506 eligible patients undergoing immune status assessment and standard echocardiography simultaneously in our center were retrospectively analyzed. The association between immune parameters and the occurrence of LVH were evaluated through univariate and multivariate logistic analysis. To develop a predictive model, we utilized four distinct modeling approaches: support vector machine (SVM), logistic regression (LR), multi-layer perceptron (MLP), and random forest (RF). In comparison to the non-LVH group, ESKD patients with LVH exhibited significantly impaired immune function, as indicated by lower cell counts of CD3+ T cells, CD4+ T cells, CD8+ T cells, and B cells. Additionally, multivariable Cox regression analysis revealed that a decrease in CD3+ T cell count was an independent risk factor for LVH, while a decrease in NK cell count was associated with the severity of LVH. The RF model demonstrated superior performance, with an average area under the curve (AUC) of 0.942. Our findings indicate a strong association between immune parameters and LVH in ESKD patients. Moreover, the RF model exhibits excellent predictive ability in identifying ESKD patients at risk of developing LVH. Based on these results, immunomodulation may represent a promising approach for preventing and treating this disease.

Prediction of the concentration of antimony in agricultural soil using data fusion, terrain attributes combined with regression kriging

Potentially toxic elements in agricultural soils are primarily derived from anthropogenic and geogenic sources. This study aims to predict and map antimony (Sb) concentration in soil using multiple regression kriging in two distinct modeling approaches, namely Sb prediction using data fusion coupled with regression kriging (scenario 1) and Sb prediction using data fusion, terrain attributes, and regression kriging (scenario 2). Cubist regression kriging (cubist_RK), conditional inference forest regression kriging (CIF_RK), extreme gradient boosting regression kriging (EGB_RK) and random forest regression kriging (RF_RK) were the modeling techniques used in the estimation of Sb concentration in agricultural soil. The validation results suggested that in scenario 1, EGB_RK was the optimal modeling approach for Sb prediction in agricultural soil with root mean square error (RMSE) = 1.31 and mean absolute error (MAE) = 0.61, bias = 0.37, and high coefficient of determination R2 = 0.81. Similarly, the EGB_RK was also the optimal modeling approach in scenario 2, with the highest R2 = 0.76, RMSE = 0.90, bias = 0.06, and MAE = 0.48 values than the other regression kriging modeling approaches. The cumulative assessment suggested that the EGB_RK in scenario 2 yielded optimal results compared to the respective modeling approach in scenario 1. The uncertainty propagated by the modeling approaches in both scenarios indicated that the degree of uncertainty during the modeling process was distributed across the study area from a low to a moderate uncertainty level. However, cubist_RK in scenario 2 exhibited some elevated spots of uncertainty levels. As a result, the combination of data fusion, terrain attributes, and regression kriging modeling approaches produces optimal results with a high R2 value, minimal errors as well as bias. Furthermore, combining terrain attributes with data fusion is promising for reducing model error, bias and yielding high-accuracy predictions.

Toward a more comprehensive modeling of sequential lineups

Sequential lineups are one of the most commonly used procedures in police departments across the USA. Although this procedure has been the target of much experimental research, there has been comparatively little work formally modeling it, especially the sequential nature of the judgments that it elicits. There are also important gaps in our understanding of how informative different types of judgments can be (binary responses vs. confidence ratings), and the severity of the inferential risks incurred when relying on different aggregate data structures. Couched in a signal detection theory (SDT) framework, the present work directly addresses these issues through a reanalysis of previously published data alongside model simulations. Model comparison results show that SDT modeling can provide elegant characterizations of extant data, despite some discrepancies across studies, which we attempt to address. Additional analyses compare the merits of sequential lineups (with and without a stopping rule) relative to showups and delineate the conditions in which distinct modeling approaches can be informative. Finally, we identify critical issues with the removal of the stopping rule from sequential lineups as an approach to capture within-subject differences and sidestep the risk of aggregation biases.

A conceptual multilevel approach to polyolefin reaction engineering

AbstractThis article gives a wide overview of different types of mathematical models that can be used to describe the polymerization of linear olefins with coordination catalysts. We expanded the conventional classification of mathematical models into micro‐, meso‐, and macroscale, to include seven modelling levels: catalysis, polymerization kinetics, thermodynamic equilibrium, particle transport phenomena, particle interactions, reactor fluid dynamics, and reactor residence time distribution. Some of these levels may coexist at the same scale, but they are better treated separately because they make use of distinct modelling approaches. How complex the models in each level need to be, as well as how many modelling levels should beimplicitlyincluded, depends on the type of application intended for the simulations. In this paper, we will argue that the proposed levels of mathematical modelling not only bring to our attention the complexity behind the simulation of laboratory‐ and industrial‐scale olefin polymerization reactors but are also useful conceptual tools to assist us to decide which levels to include and which ones to exclude when we develop simulation packages to describe olefin polymerization processes.

Crossing borders and methods: Comparing individual and social influences on energy saving in the United Arab Emirates and Germany

Research on the drivers of energy-related behaviors in buildings has gained significant interest in recent years. However, existing studies are often limited to a restricted scope of analysis (e.g., single-domain and single-cultural contexts) and rely on simplistic data analysis tools that may fail to capture complex relationships of the studied drivers of behavior. Based on a survey of university students in Abu Dhabi (UAE, n = 519) and Aachen (GER, n = 153), this study investigates the impact of demographic characteristics, personality traits, energy-related beliefs, and social perceptions on the energy conservation motivation and attempts to save energy. Two distinct modeling approaches are applied and compared: (i) traditional linear regression (often used by social scientists), and (ii) machine learning based random forest regression (often used by data scientists). When using the training and testing set from the same location, the traditional linear regression showed slightly higher predictive power and significantly higher explanatory power. In contrast, the random forest regression was more successful in predicting the participants’ attitude towards energy saving by generalizing the behavior from different locations. In either case, the most influential drivers of occupant behavior were social influence and energy conservation beliefs and ability. The Abu Dhabi and Aachen samples shared the same drivers (i.e., triggers) of behavior despite disclosing different motivation levels and attempts to conserve energy. The findings shed meaningful insights on the multi-domain nature of occupant behavior across different cultures and the premise of different data analysis approaches to capture complex relationships.

Experimental and numerical evaluation of conduction welded thermoplastic composite joints

The capability of joining two thermoplastic composite parts by welding is a key technology to reduce the weight and cost of assembled parts and enables high volume manufacturing of future aeronautical structures made of thermoplastic composite materials. However, there is not much experimental understanding of the mechanisms involving welded joint failure, and the computational tools available for the simulation of thermoset composites have not yet been completely assessed for thermoplastic materials. In this work, a numerical and experimental evaluation is performed to investigate the strength and failure behavior of conduction welded thermoplastic composite joints. A welded single lap shear joint is designed, manufactured, tested and analyzed proposing two distinct modeling approaches. A simplified modeling strategy which only accounts for damage at the weld is compared to a high-fidelity model which can take into account the physical failure mechanisms at the lamina level. The high-fidelity modeling methodology is able to predict the experimental failure mode of the investigated welded joints with high accuracy and is used to gain new insights into the key-variables that influence the strength of thermoplastic welded joints. It is also found that the joint strength is highly influenced by the failure mechanisms not only of the welded interface but also of the surrounding plies.

Matrix and agent-based modeling of threats to a diamond-backed terrapin population

Population models are important tools for evaluating human impacts and potential management approaches on declining species. However, often studies are limited by constraints of the specific modeling approach. In this study we considered the persistence of a diamond-backed terrapin (Malaclemys terrapin) population using two distinct modeling approaches. Two of the models were deterministic matrix models. Analysis of the discrete non-spatial models showed that female adult survival rate had the largest positive impact on population growth while delaying sexual maturity decreased population growth. The matrix models also demonstrated that an increase in crab traps skewed the sex ratio of the population in favor of females. The third model was a stochastic agent-based formulation that evaluated how increases in the number of crab traps and frequency of nest disturbances affected the long-term viability of diamond-backed terrapins. The spatial agent-based model revealed how terrapin mortality was highly sensitive to the proximity of traps to the primary terrapin habitat. Results from this project improve our understanding of threats to diamond-backed terrapins and can be used to guide conservation efforts.