Pasar al contenido principal

MAS ALLA DE LA MODELIZACION HABITUAL DE DATOS ESPACIALES Y TEMPORALES: UNA APROXIMACION JERARQUICA BAYESIANA

PID2022-136455NB-I00

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal para Impulsar la Investigación Científico-Técnica y su Transferencia
Subprograma Subprograma Estatal de Generación de Conocimiento
Convocatoria Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Año convocatoria 2022
Plan Estatal 2021-2023 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSITAT DE VALÈNCIA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 19
Encontrada(s) 1 página(s)

Dipòsit Digital de Documents de la UAB
Publicación Artículo científico (JournalArticle)

Spatial Bayesian distributed lag non-linear models (SB-DLNM) for small-area exposure-lag-response epidemiological modelling

  • Quijal-Zamorano, Marcos|||0000-0003-4603-5877
  • Martinez-Beneito, Miguel A.|||0000-0001-8406-8050
  • Ballester Claramunt, Joan|||0000-0003-0009-437X
  • Marí-Dell'Olmo, Marc|||0000-0003-4345-4988
Distributed lag non-linear models (DLNMs) are the reference framework for modelling lagged non-linear associations. They are usually used in large-scale multi-location studies. Attempts to study these associations in small areas either did not include the lagged non-linear effects, did not allow for geographically-varying risks or downscaled risks from larger spatial units through socioeconomic and physical meta-predictors when the estimation of the risks was not feasible due to low statistical power. Here we proposed spatial Bayesian DLNMs (SB-DLNMs) as a new framework for the estimation of reliable small-area lagged non-linear associations, and demonstrated the methodology for the case study of the temperature-mortality relationship in the 73 neighbourhoods of the city of Barcelona. We generalized location-independent DLNMs to the Bayesian framework (B-DLNMs), and extended them to SB-DLNMs by incorporating spatial models in a single-stage approach that accounts for the spatial dependence between risks. The results of the case study highlighted the benefits of incorporating the spatial component for small-area analysis. Estimates obtained from independent B-DLNMs were unstable and unreliable, particularly in neighbourhoods with very low numbers of deaths. SB-DLNMs addressed these instabilities by incorporating spatial dependencies, resulting in more plausible and coherent estimates and revealing hidden spatial patterns. In addition, the Bayesian framework enriches the range of estimates and tests that can be used in both large- and small-area studies. SB-DLNMs account for spatial structures in the risk associations across small areas. By modelling spatial differences, SB-DLNMs facilitate the direct estimation of non-linear exposure-response lagged associations at the small-area level, even in areas with as few as 19 deaths. The manuscript includes an illustrative code to reproduce the results, and to facilitate the implementation of other case studies by other researchers.
Dipòsit Digital de Documents de la UAB
Publicación Artículo científico (JournalArticle)

Comparative analysis of the yields of dicentrics and chromosomal translocations

  • Mlynarczyk, Dorota Anna|||0000-0002-7957-2567
  • Puig, Pedro|||0000-0002-6607-9642
  • Barquinero, Joan Francesc|||0000-0003-0084-5268
  • Armero, Carmen|||0000-0001-9839-6442
  • Gómez-Rubio, Virgilio|||0000-0002-4791-3072
Purpose: Chromosomal dicentrics and translocations are commonly employed as biomarkers to estimate radiation doses. The main goal of this article is to perform a comparative analysis of yields of both types of aberrations. The objective is to determine if there are relevant distinctions between both yields, allowing for a comprehensive assessment of their respective suitability and accuracy in the estimation of radiation doses. Materials and methods: The analysis involved data from a partial-radiation simulation study with the calibration data obtained through two scoring methods: conventional and PAINT modified. Subsequently, a Bayesian bivariate zero-inflated Poisson model was employed to compare the posterior marginal density of the mean of dicentrics and translocations and assess the differences between them. Results: When employing the conventional method of scoring, the findings indicate that there is no notable disparity between the yield of observed translocations and dicentrics. However, when utilizing the PAINT modified method, a notable discrepancy is observed for higher doses, indicating a relevant difference in the mean number of the two types of aberrations. Conclusions: The choice of scoring method significantly influences the analysis of radiation-induced aberrations, especially when distinguishing between complex and simple chromosomal formations. Further research and analysis are necessary to gain a deeper understanding of the factors and mechanisms impacting the formation of dicentrics and translocations.
Repositorio Digital de la UPF
Publicación Artículo científico (JournalArticle)

Spatial Bayesian distributed lag non-linear models with R-INLA

  • Quijal-Zamorano, Marcos
  • Martínez Beneito, Miguel Ángel
  • Ballester, Joan
  • Marí Dell'Olmo, Marc, 1978-
M.Q.-Z. and J.B. gratefully acknowledge funding from the European Union's Horizon 2020 and Horizon Europe research and innovation programmes under grant agreement nos 865564 (European Research Council Consolidator Grant EARLY-ADAPT, https://www.early-adapt.eu/), 101069213 (European Research Council Proof-of-Concept HHS-EWS), and 101123382 (European Research Council Proof-of-Concept FORECAST-AIR). ISGlobal authors acknowledge support from the grant CEX2018-000806-S funded by MCIN/AEI/10.13039/501100011033 and support from the Generalitat de Catalunya through the CERCA Program. M.A.M.-B. acknowledges support from Project PID2022-136455NB-I00 funded by MCIN/AEI/10.13039/501100011033/FEDER, UE.
Repositorio Digital de la UPF
Publicación Artículo científico (JournalArticle)

Spatial Bayesian distributed lag non-linear models (SB-DLNM) for small-area exposure-lag-response epidemiological modelling

  • Quijal-Zamorano, Marcos
  • Martínez Beneito, Miguel Ángel
  • Ballester, Joan
  • Marí Dell'Olmo, Marc, 1978-
Background: Distributed lag non-linear models (DLNMs) are the reference framework for modelling lagged non-linear associations. They are usually used in large-scale multi-location studies. Attempts to study these associations in small areas either did not include the lagged non-linear effects, did not allow for geographically-varying risks or downscaled risks from larger spatial units through socioeconomic and physical meta-predictors when the estimation of the risks was not feasible due to low statistical power. Methods: Here we proposed spatial Bayesian DLNMs (SB-DLNMs) as a new framework for the estimation of reliable small-area lagged non-linear associations, and demonstrated the methodology for the case study of the temperature-mortality relationship in the 73 neighbourhoods of the city of Barcelona. We generalized location-independent DLNMs to the Bayesian framework (B-DLNMs), and extended them to SB-DLNMs by incorporating spatial models in a single-stage approach that accounts for the spatial dependence between risks. Results: The results of the case study highlighted the benefits of incorporating the spatial component for small-area analysis. Estimates obtained from independent B-DLNMs were unstable and unreliable, particularly in neighbourhoods with very low numbers of deaths. SB-DLNMs addressed these instabilities by incorporating spatial dependencies, resulting in more plausible and coherent estimates and revealing hidden spatial patterns. In addition, the Bayesian framework enriches the range of estimates and tests that can be used in both large- and small-area studies. Conclusions: SB-DLNMs account for spatial structures in the risk associations across small areas. By modelling spatial differences, SB-DLNMs facilitate the direct estimation of non-linear exposure-response lagged associations at the small-area level, even in areas with as few as 19 deaths. The manuscript includes an illustrative code to reproduce the results, and to facilitate the implementation of other case studies by other researchers., M.Q-Z. and J.B. gratefully acknowledge funding from the European Union’s Horizon 2020 and Horizon Europe research and innovation programmes under grant agreement no. 865564 (European Research Council Consolidator Grant EARLY-ADAPT) [https://www.early-adapt.eu/], 101069213 (European Research Council Proof-of-Concept HHS-EWS) and 101123382 (European Research Council Proof-of-Concept FORECAST-AIR). ISGlobal authors acknowledge support from the grant CEX2018-000806-S funded by MCIN/AEI/10.13039/501100011033, and support from the Generalitat de Catalunya through the CERCA Program. M.A.M-B. acknowledges support from Project PID2022-136455NB-I00 funded by MCIN/AEI/10.13039/501100011033/FEDER, UE.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Machine learning applied to global scale species distribution models

  • Fuster-Alonso, Alba
  • Mestre-Tomás, Jorge
  • Baez, Jose Carlos
  • Pennino, Maria Grazia
  • Barber, Xavier
  • Bellido, Jose Maria
  • Conesa, David
  • Lopez-Quilez, Antonio
  • Steenbeek, Jeroen
  • Christensen, Villy
  • Coll, Marta
[EN] Species Distribution Models (SDMs) are widely used in ecology to analyze historical and future patterns of marine species distributions. Given the growing impact of climate change, predicting potential shifts in species ranges has become a key challenge. In this study, we apply Bayesian Additive Regression Trees (BART), a non-parametric machine learning algorithm, to estimate and forecast the global distribution of marine turtle species under different climate change scenarios. We model both individual species and their combined functional group, assess their historical and future habitat suitability, and examine the contribution of key environmental predictors. To evaluate BART's performance, we conduct a simulation study under two contrasting distributional scenarios: a cosmopolitan and a persistent species. We also test the sensitivity of BART to pseudo-absence data and compare its performance with MaxEnt and Generalized Additive Models (GAMs). Results indicate that BART performs slightly better overall, particularly under pseudo-absence settings, showing higher accuracy and more stable sensitivity and specificity. These findings highlight BART as a reliable alternative for long-term, global-scale species distribution modeling in marine systems., This study is a contribution to the project ProOceans (Ministerio de Ciencia e Innovacion, Proyectos de I + D + I (RETOS PID2020-118097RB-I00)). AFA received funding from the Spanish project ProOceans through an FPI grant (Ministerio de Ciencia e Innovacion, Proyectos de I + D + I (RETOSPID2020-118097RB-I00)). AFA, MC, JMT and JS acknowledge funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No. 817578 (TRIATLAS project) and No. 869300 (FutureMares). XB, DC and ALQ acknowledge support by the grant PID2022-136455NB-I00, funded by Ministerio de Ciencia, Innovacion y Universidades of Spain (MCIN/AEI/10.13039/501100011033/FEDER, UE) and the European Regional Development Fund. DC and MGP also acknowledge funding provided by the Conselleria de Educacion, Universidades y Empleo de la Generalitat Valenciana via the project CIAICO/2022/165. Authors also acknowledge institutional support of the "Severo Ochoa Center of Excellence" accreditation (CEX2024-001494-S funded by AEI 10.13039/501100011033). This project was supported by the Catalan government through the iMARES research group (2021 SGR 00435) at the Institute of Marine Sciences (ICM-CSIC) in Barcelona.This work was supported by the project SOSPEN (Spanish National Plan for Scientific and Technical Research and Innovation, 2021, PID2021-124831OA-I00).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Bayesian feedback in the framework of ecological sciences

  • Figueira, Mario
  • Barber, Xavier
  • Conesa, David
  • López-Quilez, Antonio
  • Paradinas, Iosu
  • Pennino, Maria Grazia
  • Martínez-Minaya, Joaquín|||0000-0002-1016-8734
[EN] In ecological studies, it is not uncommon to encounter scenarios where the same phenomenon (e.g., species occurrence, species abundance) is observed using two different types of samplers. For example, species data can be collected from scientific sampling with a completely random sample pattern, but also from opportunistic sampling (e.g., whale watching from commercial fishing vessels or bird watching from citizen science), where observers tend to look for particular species in areas where they expect to find them. Species Distribution Models (SDMs) are widely used tools for analysing this type of ecological data. In particular, two models are available for the aforementioned data: a geostatistical model (GM) for data collected where the sampling design is not directly related to the observations, and a preferential model (PM) for data obtained from opportunistic sampling. The integration of information from disparate sources can be addressed through the use of expert elicitation and integrated models. This paper focuses on a sequential Bayesian procedure for linking two models by updating prior distributions. The Bayesian paradigm is implemented together with the integrated nested Laplace approximation (INLA) methodology, which is an effective approach for making inference and predictions in spatial models with high performance and low computational cost. This sequential approach has been evaluated through the simulation of various scenarios and the subsequent comparison of the results from sharing information between models using a variety of criteria. The procedure has also been exemplified on a real dataset. The primary findings indicate that, in general, it is preferable to transfer information from the independent (with a completely random sampling) model to the preferential model rather than in the alternative direction. However, this depends on several factors, including the spatial range and the spatial arrangement of the sampling locations., This study is part of the ThinkInAzul program and is financed by the MCIN with funds from the European Union (NextGenerationEU-PRTR-C1711) and by the Generalitat Valenciana, Spain GVA-THINKINAZUL/2021/021. DC acknowledges Grant CIAICO/2022/165 funded by Generalitat Valenciana, Spain. XB, DC and ALQ acknowledge the support of Grant PID2022-136455NB-I00, funded by Ministerio de Ciencia, Inno-vacion y Universidades of Spain (MCIN/AEI/10.13039/501100011033/FEDER, UE) and the European Regional Development Fund. MGP also thanks the project FRESCO (PID2022-140290OB-I00) , funded by Ministerio de Ciencia, Innovacion y Universidades of Spain.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Performance of Outbreak Management Plans for Emerging Plant Diseases: The Case of Almond Leaf Scorch Caused by Xylella fastidiosa in Mainland Spain

  • Cendoya, Martina
  • Lazaro, Elena
  • López-Quilez, Antonio
  • Conesa, David
  • Vicent, Antonio
  • Navarro-Quiles, Ana
[EN] Outbreak response to quarantine pathogens and pests in the European Union (EU) is regulated by the EU Plant Health Law, but the performance of outbreak management plans in terms of their effectiveness and ef¿ciency has been quanti¿ed only to a limited extent. As a case study, the disease dynamics of almond leaf scorch, caused by Xylella fastidiosa, in the affected area of Alicante, Spain, were approximated using an individual-based spatial epidemiological model. The emergence of this outbreak was dated based on phylogenetic studies, and of¿cial surveys were used to delimit the current extent of the disease. Different survey strategies and disease control measures were compared to determine their effectiveness and ef¿ciency for outbreak management in relation to a baseline scenario without interventions. One-step and two-step survey approaches were compared with different con¿dence levels, buffer zone sizes, and eradication radii, including those set by the EU legislation for X. fastidiosa. The effect of disease control interventions was also considered by decreasing the transmission rate in the buffer zone. All outbreak management plans reduced the number of infected trees (effectiveness), but large differences were observed in the number of susceptible trees not eradicated (ef¿ciency). The two-step survey approach, high con¿dence level, and the reduction in the transmission rate increased the ef¿ciency. Only the outbreak management plans with the two-step survey approach removed infected trees completely, but they required greater survey efforts. Although control measures reduced disease spread, surveillance was the key factor in the effectiveness and ef¿ciency of the outbreak management plan, Support was provided by Horizon Europe (grant 101060593; BeXyl, Beyond Xylella, Integrated Management Strategies for Mitigating Xylella fastidiosa Impact in Europe ), Ministerio de Ciencia e Innovación (project PID2022-136455NBI00 awarded to D. Conesa and A. López-Quílez), the European Regional Development Fund, and Generalitat Valenciana (grant CIAICO/2022/165 awarded to D. Conesa).
Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
Publicación Artículo científico (JournalArticle)

A proposal for homoskedastic modeling with conditional auto-regressive distributions

  • Martínez-Beneito, Miguel A.
  • Adin Urtasun, Aritz
  • Goicoa Mangado, Tomás
  • Ugarte Martínez, María Dolores
Conditional auto-regressive (CAR) distributions are widely used to deal with spatial dependence in the geographic analysis of areal data. These distributions establish multivariate dependence networks by defining conditional relationships between neighboring units, resulting in positive dependence among nearby observations. Despite their practical convenience and well-founded principles, the conditional nature of CAR distributions can lead to undesirable marginal properties, such as inherent heteroskedasticity assumptions that may significantly impact the posterior distributions. In this paper, we highlight the variance issues associated with CAR distributions, particularly focusing on edge effects and issues related to the region's geometry. We show that edge effects may be more pronounced and widespread in disease mapping studies than previously anticipated. To address these heteroskedasticity concerns, we introduce a new conditional autoregressive distribution designed to mitigate these problems. We demonstrate how this distribution effectively diminishes the practical issues identified in earlier models., This work was supported by projects PID2020-113125RB-I00/MCIN/AEI/10.13039/501100011033, PID2022-136455NB-I00/MCIN/AEI/10.13039/501100011033, and PID2024-155382OB-I00/MICIU/AEI/10.13039/501100011033 funded by the Ministry of Science, Innovation and Universities of Spain andFEDER (EU), and Grant CIAICO/2022/165 funded by Generalitat Valenciana.
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Bayesian joint longitudinal models for assessing the exploitation rates of sardine stock in the Mediterranean Sea

  • Calvo, Gabriel
  • Armero, Carmen
  • Spezia, Luigi
  • Pennino, Maria Grazia
The European sardine is a pelagic species of great ecological importance for the conservation of the Mediterranean Sea as well as economic importance for the Mediterranean countries. Its fishing has suffered a significant decline in recent years due to various economic, cultural and ecological reasons. This paper focuses on the evolution of sardine catches in the Mediterranean Sea from 1985 to 2018 according to the fishing Mediterranean country and the type of fishing practised, artisanal and industrial. We propose three Bayesian longitudinal linear mixed models to assess differences in the temporal evolution of artisanal and industrial fisheries between and within countries. Overall results confirm that Mediterranean fishery time series are highly diverse along their dynamics and this heterogeneity is persistent throughout the time. Furthermore, our results highlight a positive correlation between artisanal and industrial fishing. Finally, the study observes a consistent decreasing time trend in the quantity of fish landings. Although the causes of this feature could be also linked to economic motivations (such as a reduction in demand or the reorientation of fleets towards more commercially beneficial species), it may indicate a potential risk to the stock of this species in the Mediterranean Sea., Gabriel Calvo’s research was partially funded by the ONCE Foundation, the Universia Foundation, and the Spanish Ministry of Education and Professional Training, grant FPU18/03101. Carmen Armero and Gabriel Calvo gratefully acknowledge the Ministerio de Ciencia e Innovación (Spain) for financial support (jointly financed by the European Regional Development Fund) via Research Grant PID2022-136455NB-I00. Luigi Spezia’s research was funded by the Scottish Government’s Rural and Environment Science and Analytical Services Division., Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Accounting for spatio-temporal and sampling dependence in survey and CPUE biomass indices: simulation and Bayesian modeling framework

  • Fuster-Alonso, Alba
  • Conesa, David
  • Cousido-Rocha, Marta
  • Izquierdo, Francisco
  • Paradinas, Iosu
  • Cerviño, Santiago
  • Pennino, Maria Grazia
12 pages, 6 figures, 1 table, supplementary data https://doi.org/10.1093/icesjms/fsae056, Estimating changes in the biomass of a fish stock is crucial for successful management. However, fishery assessment may be affected by the quality of the inputs used in stock assessment models. Survey biomass indices derived from fishery-independent and catch per unit effort (CPUE) biomass indices derived from fishery-dependent data are key inputs for model calibration. These indices have biases that could compromise the accuracy of the stock assessment models results. Therefore, there are plenty proposed methods to standardize survey or CPUE biomass data. From simpler models like generalized linear models (GLMs) to more complex models that take into account spatio-temporal correlation, like geostatistical models, and sampling dependence, like marked point processes. But many of them do not consider the underlying spatio-temporal or sampling dependence of these data. Hence, the goal of the study is to present a spatio-temporal simulation and Bayesian modeling framework to assess the impact of applying models that do not consider spatio-temporal and sampling dependence. Results indicate that geostatistical models and marked point processes achieve the lowest measures of error. Hence, to capture the underlying spatio-temporal process of the survey and CPUE biomass indices and data sampling preferentiality, it is essential to apply models that consider the spatio-temporal and sampling dependence, This study is a contribution to the project IMPRESS (RTI2018-099868-B-I00), ERDF, Ministry of Science, Innovation and Universities - State Research Agency and also of the projects financed by the European Union-Next Generation EU. Componente 3. Inversión 7. CONVENIO ENTRE EL MINISTERIO DE AGRICULTURA, PESCA, Y ALIMENTACIÓN Y LA AGENCIA ESTATAL CONSEJO SUPERIOR DE INVESTIGACIONES CIENTÍFICAS M.P. -A TRAVÉS DEL INSTITUTO ESPAÑOL DE OCEANOGRAFÍA- PARA IMPULSAR LA INVESTIGACIÓN PESQUERA COMO BASE PARA LA GESTIÓN PESQUERA SOSTENIBLE. Eje4, FishClim: Conocimiento científico para la adaptación al cambio climático del sector pesquero español and Eje6, Math4Fish: Nuevas herramientas para el modelado matemático en el asesoramiento científico de pesquerías españolas. Also, we were supported by the GAIN [Agencia Gallega de Innovación] - Xunta de Galicia, GRC-MERVEX (IN607A 2022/04). David Conesa was supported by grant PID2022-136455NB-I00, funded by Ministerio de Ciencia, Innovación y Universidades of Spain (MCIN/AEI/10.13039/501100011033/FEDER, UE) and the European Regional Development Fund. David Conesa was also funded by grant CIAICO/2022/165 funded by Generalitat Valenciana. Alba Fuster-Alonso received funding from the Spanish project ProOceans (Ministerio de Ciencia e Innovación, Proyectos de I + D + I (RETOSPID2020-118097RB-I00)). A.F.-A. was supported by Ministerio de Ciencia e Innovación, Grant no. PRE2021-099287 from the project ProOeans (PID2020-118097RB-I00). Alba Fuster-Alonso was supported by the Barcelona municipal government through the iMARES research group at the Institute of Marine Sciences (ICM-CSIC) in Barcelona, With the institutional support of the ‘Severo Ochoa Centre of Excellence’ accreditation (CEX2019-000928-S), Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Bayesian feedback in the framework of ecological sciences

  • Figueira, Mario
  • Barber, Xavier
  • Conesa, David
  • López-Quílez, Antonio
  • Martínez-Minaya, Joaquín
  • Paradinas, Iosu
  • Pennino, Maria Grazia
In ecological studies, it is not uncommon to encounter scenarios where the same phenomenon (e.g., species occurrence, species abundance) is observed using two different types of samplers. For example, species data can be collected from scientific sampling with a completely random sample pattern, but also from opportunistic sampling (e.g., whale watching from commercial fishing vessels or bird watching from citizen science), where observers tend to look for particular species in areas where they expect to find them. Species Distribution Models (SDMs) are widely used tools for analysing this type of ecological data. In particular, two models are available for the aforementioned data: a geostatistical model (GM) for data collected where the sampling design is not directly related to the observations, and a preferential model (PM) for data obtained from opportunistic sampling. The integration of information from disparate sources can be addressed through the use of expert elicitation and integrated models. This paper focuses on a sequential Bayesian procedure for linking two models by updating prior distributions. The Bayesian paradigm is implemented together with the integrated nested Laplace approximation (INLA) methodology, which is an effective approach for making inference and predictions in spatial models with high performance and low computational cost. This sequential approach has been evaluated through the simulation of various scenarios and the subsequent comparison of the results from sharing information between models using a variety of criteria. The procedure has also been exemplified on a real dataset. The primary findings indicate that, in general, it is preferable to transfer information from the independent (with a completely random sampling) model to the preferential model rather than in the alternative direction. However, this depends on several factors, including the spatial range and the spatial arrangement of the sampling locations., This study is part of the ThinkInAzul program and is financed by the MCIN with funds from the European Union (NextGenerationEU-PRTR-C1711) and by the Generalitat Valenciana, Spain GVA-THINKINAZUL/ 2021/021. DC acknowledges Grant CIAICO/2022/165 funded by Generalitat Valenciana, Spain. XB, DC and ALQ acknowledge the support of Grant PID2022-136455NB-I00, funded by Ministerio de Ciencia, Innovación Universidades of Spain (MCIN/AEI/10.13039/501100011033/ FEDER, UE) and the European Regional Development Fund . MGP also thanks the project FRESCO (PID2022-140290OB-I00), funded by Ministerio de Ciencia, Innovación Universidades of Spain ., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Assessment of bivariate relationships between spatial patterns: Revisiting the global and local L bivariate indices for wildlife management and conservation

  • Illanas, Sonia
  • Gómez-Rubio, Virgilio
  • Vicente, Joaquín
  • Acevedo, Pelayo
In wildlife management and conservation, the spatial relationship between two species (e.g. abundance patterns) is often studied through overlapping local information or implementing clustering analysis. However, conventional approaches overlook processes operating at larger spatial scales than the observation unit. Analyses that consider not only a location but also its neighbors, such as the Global L bivariate index (GL) and its Local (LL) measure, can be more appropriate for addressing spatial relationships between two species. The current formulation considers a unique connectivity matrix to define neighbors’ structure, assuming both species exhibit the same spatial dependence even when they may differ. We revisited the Global (GL′) and Local (LL′) indices to incorporate specific connectivity matrices to enhance wildlife applicability. We studied the behavior of the revisited indices through a simulation and a case study of four wildlife species with contrasting movement capabilities. The simulation study revealed the indices’ intrinsic characteristics, including smaller GL′ values and smoother LL′ patterns as the number of neighbors increases. The case study suggested imbalanced coexistence and divergence areas for the species studied, providing meaningful scenarios for wildlife management. The application of species-specific connectivity matrices facilitates the consideration of spatial use characteristics, offering a more reliable understanding of ecological relationships, particularly when the observation unit is smaller than the species’ home range. These findings are crucial for determining accurate and ecologically meaningful connectivity matrices. Overall, the revisited indices offer high flexibility and a comprehensive description of the spatial relationships between ecological indicators., This research was supported by the Spanish Ministry of Science and Innovation thanks to the 2020 PhD contract (PRE2020-095091) associated to I+D+i HAWIPO Project (PID2019-111699RB-I00), the I+D+I AGROBOAR Project (PID2022-142919OB-100) funded by MCIU/AEI/10.13039/501100011033/y – FEDER – EU, and the IBERCONEJO LIFE Project funded by the European Union. Our research group is partly supported by 2022-GRIN-34227 grant, funded by the University of Castilla-La Mancha and FEDER. Virgilio Gómez-Rubio was supported by the Consejería de Educación, Cultura y Deportes (JCCM, Spain) and Fondo Europeo de Desarrollo Regional thanks to a grant (SBPLY/21/180501/000241), by the Spanish Ministry of Science and Innovation (PID2022-136455NB-I00), and by the Dirección General de Ciencia e Investigación (Generalitat Valenciana; CIAICO/2022/165)., Open Access funding provided by Universidad de Castilla-La Mancha., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Machine learning applied to global scale species distribution models

  • Fuster-Alonso, Alba
  • Mestre-Tomás, Jorge
  • Báez, José Carlos
  • Pennino, Maria Grazia
  • Barber, Xavier
  • Bellido, José M.
  • Conesa, David
  • López-Quílez, Antonio
  • Steenbeek, Jeroen
  • Christensen, Villy
  • Coll, Marta
[Data availability] The datasets generated and/or analysed during the current study are available in the Global Biodiversity Information Facility (GBIF) repository, https://www.gbif.org/es/. DOIs are available in the supplementary material., Species Distribution Models (SDMs) are widely used in ecology to analyze historical and future patterns of marine species distributions. Given the growing impact of climate change, predicting potential shifts in species ranges has become a key challenge. In this study, we apply Bayesian Additive Regression Trees (BART), a non-parametric machine learning algorithm, to estimate and forecast the global distribution of marine turtle species under different climate change scenarios. We model both individual species and their combined functional group, assess their historical and future habitat suitability, and examine the contribution of key environmental predictors. To evaluate BART’s performance, we conduct a simulation study under two contrasting distributional scenarios: a cosmopolitan and a persistent species. We also test the sensitivity of BART to pseudo-absence data and compare its performance with MaxEnt and Generalized Additive Models (GAMs). Results indicate that BART performs slightly better overall, particularly under pseudo-absence settings, showing higher accuracy and more stable sensitivity and specificity. These findings highlight BART as a reliable alternative for long-term, global-scale species distribution modeling in marine systems., This work was supported by the project SOSPEN (Spanish National Plan for Scientific and Technical Research and Innovation, 2021, PID2021-124831OA-I00)., This study is a contribution to the project ProOceans (Ministerio de Ciencia e Innovación, Proyectos de I + D + I (RETOS PID2020-118097RB-I00)). AFA received funding from the Spanish project ProOceans through an FPI grant (Ministerio de Ciencia e Innovación, Proyectos de I + D + I (RETOSPID2020-118097RB-I00)). AFA, MC, JMT and JS acknowledge funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 817578 (TRIATLAS project) and No. 869300 (FutureMares). XB, DC and ALQ acknowledge support by the grant PID2022-136455NB-I00, funded by Ministerio de Ciencia, Innovación y Universidades of Spain (MCIN/AEI/10.13039/501100011033/FEDER, UE) and the European Regional Development Fund. DC and MGP also acknowledge funding provided by the Conselleria de Educación, Universidades y Empleo de la Generalitat Valenciana via the project CIAICO/2022/165. Authors also acknowledge institutional support of the “Severo Ochoa Center of Excellence” accreditation (CEX2024-001494-S funded by AEI 10.13039/501100011033). This project was supported by the Catalan government through the iMARES research group (2021 SGR 00435) at the Institute of Marine Sciences (ICM-CSIC) in Barcelona, Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Bayes factors for longitudinal model assessment via power posteriors

  • Calvo, Gabriel
  • Armero, Carmen
  • Spezia, Luigi
  • Pennino, Maria Grazia
Bayes factor, defined as the ratio of the marginal likelihood functions of two competing models, is the natural Bayesian procedure for model selection. Marginal likelihoods are usually computationally demanding and complex. This scenario is particularly cumbersome in linear mixed models (LMMs) because the marginal likelihood functions involve integrals of large dimensions determined by the number of parameters and the number of random effects, which in turn increase with the number of individuals in the sample. Power posterior is an attractive proposal in the context of the Markov chain Monte Carlo algorithms that allows expressing marginal likelihoods as one-dimensional integrals over the unit range. This paper explores the use of power posteriors in LMMs and discusses its behavior through three simulation studies and a real data set on European sardine landings in the Mediterranean Sea., Gabriel Calvo’s research was partially funded by the ONCE Foundation, the Universia Foundation, and the Spanish Ministry of Education and Professional Training, grant FPU18/03101. This paper is part of the project PID2022-136455NB-I00, funded by Ministerio de Ciencia, Innovación y Universidades of Spain (MCIN/AEI/10.13039/ 501100011033/FEDER, UE) and the European Regional Development Fund. MGP’s research was funded by the project IMPRESS (RTI2018-099868-B-I00), ERDF, Ministry of Science, Innovation and Universities - State Research Agency. Luigi Spezia’s research was funded by the Scottish Government’s Rural and Environment Science and Analytical Services Division., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Mapping Escolar (Lepidocybium flavobrunneum) in Motion: Oceanographic Forces Shaping Its Habitat in the Southwestern South Atlantic, With Insights From Fishers' Perceptions

  • Rodrigues, Lucas
  • Pennino, Maria Grazia
  • Kinas, Paul
  • Gerhard Conesa, David
  • Fuster-Alonso, Alba
  • Domingo, Andrés
  • Forselledo, Rodrigo
  • Mas, Federico
  • Mourato, Bruno
  • Sant'Ana, Rodrigo
  • Santos Thykjaer, Vinni
  • Marquez Marquez, Raquel Virginia
  • Ventura Alves, Abner
  • Cabanillas-Torpoco, Mariano
  • Schroeder, Rafael
  • Cardoso, Luis Gustavo
[Data Availability Statement] The data that support the findings of this study are available from the corresponding author upon reasonable request., Lepidocybium flavobrunneum [Smith, 1843], commonly known as escolar, is a large pelagic species, important for global and local fisheries, particularly in the southwestern South Atlantic Ocean (SWAO), where it constitutes a significant portion of the catch. However, despite its important socioeconomic role, it remains less studied compared to other large pelagic species. Enhancing research efforts is therefore key to better understanding the ecological processes that influence its spatial and temporal distribution. To address this gap, our study applied species distribution models (SDMs) to escolar's occurrence and abundance using longline fleet data from south Brazil and Uruguay. Estimation and prediction were conducted at the Bayesian framework using the integrated nested Laplace approximation (INLA) methodology combined with the stochastic partial differential equation (SPDE) approach. Our findings reveal a strong correlation between warmer sea surface temperatures and increased escolar occurrence and abundance in the SWAO, particularly during the austral summer, in contrast to cooler-season patterns observed elsewhere. Escolar's distribution is influenced by the southward and westward movement of seawater driven by the Brazil Current, and its preference for depths greater than 2000 m reflects its oceanic nature. Predicted areas of higher abundance around the continental slope (between 33° and 37° S) and the Rio Grande Rise are linked to the dynamic interaction of warm and cold currents and nutrient-rich waters. Additionally, escolar's sensitivity to low-light conditions may explain reduced catches during moonlit nights. A complementary analysis assessing the synergistic influence of sea surface temperature and lunar illumination on escolar occurrence indicated that, under high temperatures (> 22 °C), occurrence is further enhanced during dark nights. Furthermore, semi-structured interviews with vessel captains revealed that 90% identified temperature (via open-ended responses) as a key environmental factor driving escolar catches, and 80% of them attributed a strong influence (greater than average catch rates) to temperature, summer season and new moon phase. This study provides new insights into escolar's seasonal and spatial patterns, emphasizing the impact of oceanographic features and light conditions on its distribution. It also highlights the need for further research into the mechanisms driving escolar populations., We would like to express our gratitude to everyone involved in data collection for this study, including those responsible for filling out the logbooks and the onboard observers, whose dedication ensured the quality of the information. Your commitment and hard work were essential to the success of this research. Special thanks go to the fishers and crews who kindly allowed presence onboard, as well as to the coordination and logistics teams, whose support was invaluable in facilitating the field sampling efforts. LdSR was supported by a Doctoral (88887.374239/2019-00) and science-without-boarders (88887.717506/2022-00) grants (Programa de Doutorado-sanduíche no Exterior [PDSE]) from the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior—CAPES. DC thanks support by Grant PID2022-136455NB-I00 (funded by Ministerio de Ciencia, Innovación y Universidades of Spain; MCIN/AEI/https://doi.org/10.13039/501100011033/FEDER, UE), Grant CIAICO/2022/165 (funded by Generalitat Valenciana), and Grant RED2022-134202-T (also funded by Ministerio de Ciencia, Innovación y Universidades of Spain). This paper is a contribution of LADIPP, PPGOB, IO, FURG., This work was supported by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior, 10.13039/501100002322, 88887.374239/2019-00, 88887.717506/2022-00; Ministerio de Ciencia, Innovación y Universidades of Spain, PID2022-136455NB-I00, RED2022-134202-T; and Generalitat Valenciana, 10.13039/501100003359, CIAICO/2022/165., Peer reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Bayesian estimation for conditional probabilities associated to directed acyclic graphs: study of hospitalization of severe influenza cases

  • Acosta Argueta, Lesly María|||0000-0002-5859-2120
  • Armero, Carmen
This paper presents a Bayesian framework to estimate joint, conditional, and marginal probabilities in directed acyclic graphs to study the progression of hospitalized patients with confrmed severe infuenza. Using data from the PIDIRAC retrospective cohort in Catalonia, we model patient pathways from admission to discharge, death, or transfer. Transition probabilities are estimated using a Bayesian Dirichlet-multinomial approach, while posterior distributions for absorbing states or inverse probabilities are assessed via simulation. Bayesian methodology quantifes uncertainty through posterior distributions, offering insights into disease progression and in improving hospital planning. These fndings support more effective patient management and informed decision making during seasonal infuenza outbreaks., The authors are very grateful to two anonymous reviewers, whose valuable comments and suggestions have contributed signifcantly to improving the quality of the paper. This paper is partially supported by the project PID2023-148158OB-I00, funded by Ministerio de Ciencia, Innovación y Universidades (Spain) and the project PID2022- 136455NB-I00, funded by Ministerio de Ciencia, Innovación y Universidades (Spain) (MCIN/AEI/10.13039 /501100011033/FEDER, UE) and the European Regional Development Fund., Peer Reviewed
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Spatial Bayesian distributed lag non-linear models (SB-DLNM) for small-area exposure-lag-response epidemiological modelling

  • Quijal-Zamorano, Marcos
  • Martínez Beneito, Miguel Ángel
  • Ballester, Joan
  • Marí Dell'Olmo, Marc, 1978-
Background: Distributed lag non-linear models (DLNMs) are the reference framework for modelling lagged non-linear associations. They are usually used in large-scale multi-location studies. Attempts to study these associations in small areas either did not include the lagged non-linear effects, did not allow for geographically-varying risks or downscaled risks from larger spatial units through socioeconomic and physical meta-predictors when the estimation of the risks was not feasible due to low statistical power. Methods: Here we proposed spatial Bayesian DLNMs (SB-DLNMs) as a new framework for the estimation of reliable small-area lagged non-linear associations, and demonstrated the methodology for the case study of the temperature-mortality relationship in the 73 neighbourhoods of the city of Barcelona. We generalized location-independent DLNMs to the Bayesian framework (B-DLNMs), and extended them to SB-DLNMs by incorporating spatial models in a single-stage approach that accounts for the spatial dependence between risks. Results: The results of the case study highlighted the benefits of incorporating the spatial component for small-area analysis. Estimates obtained from independent B-DLNMs were unstable and unreliable, particularly in neighbourhoods with very low numbers of deaths. SB-DLNMs addressed these instabilities by incorporating spatial dependencies, resulting in more plausible and coherent estimates and revealing hidden spatial patterns. In addition, the Bayesian framework enriches the range of estimates and tests that can be used in both large- and small-area studies. Conclusions: SB-DLNMs account for spatial structures in the risk associations across small areas. By modelling spatial differences, SB-DLNMs facilitate the direct estimation of non-linear exposure-response lagged associations at the small-area level, even in areas with as few as 19 deaths. The manuscript includes an illustrative code to reproduce the results, and to facilitate the implementation of other case studies by other researchers., M.Q-Z. and J.B. gratefully acknowledge funding from the European Union’s Horizon 2020 and Horizon Europe research and innovation programmes under grant agreement no. 865564 (European Research Council Consolidator Grant EARLY-ADAPT) [https://www.early-adapt.eu/], 101069213 (European Research Council Proof-of-Concept HHS-EWS) and 101123382 (European Research Council Proof-of-Concept FORECAST-AIR). ISGlobal authors acknowledge support from the grant CEX2018-000806-S funded by MCIN/AEI/10.13039/501100011033, and support from the Generalitat de Catalunya through the CERCA Program. M.A.M-B. acknowledges support from Project PID2022-136455NB-I00 funded by MCIN/AEI/10.13039/501100011033/FEDER, UE.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Bayesian estimation for conditional probabilities associated to directed acyclic graphs: study of hospitalization of severe influenza cases

  • Acosta Argueta, Lesly María
  • Armero, Carmen
This paper presents a Bayesian framework to estimate joint, conditional, and marginal probabilities in directed acyclic graphs to study the progression of hospitalized patients with confrmed severe infuenza. Using data from the PIDIRAC retrospective cohort in Catalonia, we model patient pathways from admission to discharge, death, or transfer. Transition probabilities are estimated using a Bayesian Dirichlet-multinomial approach, while posterior distributions for absorbing states or inverse probabilities are assessed via simulation. Bayesian methodology quantifes uncertainty through posterior distributions, offering insights into disease progression and in improving hospital planning. These fndings support more effective patient management and informed decision making during seasonal infuenza outbreaks., The authors are very grateful to two anonymous reviewers, whose valuable comments and suggestions have contributed signifcantly to improving the quality of the paper. This paper is partially supported by the project PID2023-148158OB-I00, funded by Ministerio de Ciencia, Innovación y Universidades (Spain) and the project PID2022- 136455NB-I00, funded by Ministerio de Ciencia, Innovación y Universidades (Spain) (MCIN/AEI/10.13039 /501100011033/FEDER, UE) and the European Regional Development Fund., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Spatial Bayesian distributed lag non-linear models with R-INLA

  • Quijal-Zamorano, Marcos
  • Martínez Beneito, Miguel Ángel
  • Ballester, Joan
  • Marí Dell'Olmo, Marc, 1978-
M.Q.-Z. and J.B. gratefully acknowledge funding from the European Union's Horizon 2020 and Horizon Europe research and innovation programmes under grant agreement nos 865564 (European Research Council Consolidator Grant EARLY-ADAPT, https://www.early-adapt.eu/), 101069213 (European Research Council Proof-of-Concept HHS-EWS), and 101123382 (European Research Council Proof-of-Concept FORECAST-AIR). ISGlobal authors acknowledge support from the grant CEX2018-000806-S funded by MCIN/AEI/10.13039/501100011033 and support from the Generalitat de Catalunya through the CERCA Program. M.A.M.-B. acknowledges support from Project PID2022-136455NB-I00 funded by MCIN/AEI/10.13039/501100011033/FEDER, UE.