Anthony R. Charsley, Arnaud Grüss, Nokuthaba Sibanda, Shannan K. Crow, Owen F. Anderson, Ashley A. Rowden, Simon D. Hoyle, David D. Bowden
{"title":"Integrating Presence-Only Data Into Spatio-Temporal Models to Support Fisheries Assessments and Management in Freshwater and Marine Environments","authors":"Anthony R. Charsley, Arnaud Grüss, Nokuthaba Sibanda, Shannan K. Crow, Owen F. Anderson, Ashley A. Rowden, Simon D. Hoyle, David D. Bowden","doi":"10.1111/faf.12907","DOIUrl":null,"url":null,"abstract":"<p>Spatio-temporal species distribution models can support fisheries assessments and management in marine and freshwater environments. However, the high costs of structured surveys often limit the spatio-temporal coverage of the data available for modelling. To address this issue, we present a spatio-temporal modelling approach integrating structured survey data with unstructured presence-only data, which have greater spatio-temporal coverage than structured data, but are often disregarded in fisheries research. Data integration is achieved by generating pseudo-absences for the presence-only data and estimating spatially varying catchability for all data sources relative to the structured dataset. We consider a freshwater application, building longfin eel (<i>Anguilla dieffenbachii</i>, Anguillidae) spatio-temporal distribution models for the Taranaki region, New Zealand, and a marine application, building spatial density models for the vulnerable marine ecosystem indicator taxon Demospongiae in the South Pacific Ocean. We also conduct a simulation experiment to investigate the impacts of using pseudo-absences that do not reflect true absence patterns in our modelling framework. By integrating unstructured presence-only data, our approach improves the spatio-temporal coverage of the data available for modelling. Our applications provide results consistent with previous modelling studies but also offer new insights into the distribution and density patterns of longfin eel and Demospongiae. The simulation experiment found greater error and poorer uncertainty characterisation in models that mis-specified true absence patterns. We recommend assessing spatial structure in presence-only data and generating spatially structured pseudo-absences that match this structure. Our approach has many potential applications, such as providing enhanced information to assist fisheries in assessments and management.</p>","PeriodicalId":169,"journal":{"name":"Fish and Fisheries","volume":"26 4","pages":"699-716"},"PeriodicalIF":6.1000,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/faf.12907","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Fish and Fisheries","FirstCategoryId":"97","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1111/faf.12907","RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"FISHERIES","Score":null,"Total":0}
引用次数: 0
Abstract
Spatio-temporal species distribution models can support fisheries assessments and management in marine and freshwater environments. However, the high costs of structured surveys often limit the spatio-temporal coverage of the data available for modelling. To address this issue, we present a spatio-temporal modelling approach integrating structured survey data with unstructured presence-only data, which have greater spatio-temporal coverage than structured data, but are often disregarded in fisheries research. Data integration is achieved by generating pseudo-absences for the presence-only data and estimating spatially varying catchability for all data sources relative to the structured dataset. We consider a freshwater application, building longfin eel (Anguilla dieffenbachii, Anguillidae) spatio-temporal distribution models for the Taranaki region, New Zealand, and a marine application, building spatial density models for the vulnerable marine ecosystem indicator taxon Demospongiae in the South Pacific Ocean. We also conduct a simulation experiment to investigate the impacts of using pseudo-absences that do not reflect true absence patterns in our modelling framework. By integrating unstructured presence-only data, our approach improves the spatio-temporal coverage of the data available for modelling. Our applications provide results consistent with previous modelling studies but also offer new insights into the distribution and density patterns of longfin eel and Demospongiae. The simulation experiment found greater error and poorer uncertainty characterisation in models that mis-specified true absence patterns. We recommend assessing spatial structure in presence-only data and generating spatially structured pseudo-absences that match this structure. Our approach has many potential applications, such as providing enhanced information to assist fisheries in assessments and management.
期刊介绍:
Fish and Fisheries adopts a broad, interdisciplinary approach to the subject of fish biology and fisheries. It draws contributions in the form of major synoptic papers and syntheses or meta-analyses that lay out new approaches, re-examine existing findings, methods or theory, and discuss papers and commentaries from diverse areas. Focal areas include fish palaeontology, molecular biology and ecology, genetics, biochemistry, physiology, ecology, behaviour, evolutionary studies, conservation, assessment, population dynamics, mathematical modelling, ecosystem analysis and the social, economic and policy aspects of fisheries where they are grounded in a scientific approach. A paper in Fish and Fisheries must draw upon all key elements of the existing literature on a topic, normally have a broad geographic and/or taxonomic scope, and provide general points which make it compelling to a wide range of readers whatever their geographical location. So, in short, we aim to publish articles that make syntheses of old or synoptic, long-term or spatially widespread data, introduce or consolidate fresh concepts or theory, or, in the Ghoti section, briefly justify preliminary, new synoptic ideas. Please note that authors of submissions not meeting this mandate will be directed to the appropriate primary literature.