Connectivity modelling in conservation science: a comparative evaluation
Abstract Landscape connectivity, the extent to which a landscape facilitates the flow of ecological processes such as organism movement, has grown to become a central focus of applied ecology and conservation science. Several computational algorithms have been developed to understand and map connect...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2022-10-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-022-20370-w |
| _version_ | 1811194922679664640 |
|---|---|
| author | Siddharth Unnithan Kumar Samuel A. Cushman |
| author_facet | Siddharth Unnithan Kumar Samuel A. Cushman |
| author_sort | Siddharth Unnithan Kumar |
| collection | DOAJ |
| description | Abstract Landscape connectivity, the extent to which a landscape facilitates the flow of ecological processes such as organism movement, has grown to become a central focus of applied ecology and conservation science. Several computational algorithms have been developed to understand and map connectivity, and many studies have validated their predictions using empirical data. Yet at present, there is no published comparative analysis which uses a comprehensive simulation framework to measure the accuracy and performance of the dominant methods in connectivity modelling. Given the widespread usage of such models in spatial ecology and conservation science, a thorough evaluation of their predictive abilities using simulation techniques is essential for guiding their appropriate and effective application across different contexts. In this paper, we address this by using the individual-based movement model Pathwalker to simulate different connectivity scenarios generated from a wide range of possible movement behaviours and spatial complexities. With this simulated data, we test the predictive abilities of three major connectivity models: factorial least-cost paths, resistant kernels, and Circuitscape. Our study shows the latter two of these three models to consistently perform most accurately in nearly all cases, with their abilities varying substantially in different contexts. For the majority of conservation applications, we infer resistant kernels to be the most appropriate model, except for when the movement is strongly directed towards a known location. We conclude this paper with a review and interdisciplinary discussion of the current limitations and possible future developments of connectivity modelling. |
| first_indexed | 2024-04-12T00:35:21Z |
| format | Article |
| id | doaj.art-84dd202a07774e9d9f1532ada6037702 |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-04-12T00:35:21Z |
| publishDate | 2022-10-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-84dd202a07774e9d9f1532ada60377022022-12-22T03:55:11ZengNature PortfolioScientific Reports2045-23222022-10-0112111210.1038/s41598-022-20370-wConnectivity modelling in conservation science: a comparative evaluationSiddharth Unnithan Kumar0Samuel A. Cushman1Mathematical Institute, University of OxfordWildlife Conservation Research Unit (WildCRU), Department of Zoology, University of OxfordAbstract Landscape connectivity, the extent to which a landscape facilitates the flow of ecological processes such as organism movement, has grown to become a central focus of applied ecology and conservation science. Several computational algorithms have been developed to understand and map connectivity, and many studies have validated their predictions using empirical data. Yet at present, there is no published comparative analysis which uses a comprehensive simulation framework to measure the accuracy and performance of the dominant methods in connectivity modelling. Given the widespread usage of such models in spatial ecology and conservation science, a thorough evaluation of their predictive abilities using simulation techniques is essential for guiding their appropriate and effective application across different contexts. In this paper, we address this by using the individual-based movement model Pathwalker to simulate different connectivity scenarios generated from a wide range of possible movement behaviours and spatial complexities. With this simulated data, we test the predictive abilities of three major connectivity models: factorial least-cost paths, resistant kernels, and Circuitscape. Our study shows the latter two of these three models to consistently perform most accurately in nearly all cases, with their abilities varying substantially in different contexts. For the majority of conservation applications, we infer resistant kernels to be the most appropriate model, except for when the movement is strongly directed towards a known location. We conclude this paper with a review and interdisciplinary discussion of the current limitations and possible future developments of connectivity modelling.https://doi.org/10.1038/s41598-022-20370-w |
| spellingShingle | Siddharth Unnithan Kumar Samuel A. Cushman Connectivity modelling in conservation science: a comparative evaluation Scientific Reports |
| title | Connectivity modelling in conservation science: a comparative evaluation |
| title_full | Connectivity modelling in conservation science: a comparative evaluation |
| title_fullStr | Connectivity modelling in conservation science: a comparative evaluation |
| title_full_unstemmed | Connectivity modelling in conservation science: a comparative evaluation |
| title_short | Connectivity modelling in conservation science: a comparative evaluation |
| title_sort | connectivity modelling in conservation science a comparative evaluation |
| url | https://doi.org/10.1038/s41598-022-20370-w |
| work_keys_str_mv | AT siddharthunnithankumar connectivitymodellinginconservationscienceacomparativeevaluation AT samuelacushman connectivitymodellinginconservationscienceacomparativeevaluation |